#avr | Logs for 2011-12-11

Back
[02:34:38] <mapee> morning
[02:35:20] <lifeeth> morning
[03:46:46] * inflex ponders submitting his first brushless ESC design
[04:07:06] <Igor2_> hi all
[04:07:40] <Igor2_> i'd like to get some feedback about my approach of programming AVR fuse bits: http://repo.hu/projects/shscripts/avrfuse/
[04:43:09] <grummund> Igor2_: do you know that fuse settings can be placed in code to appear in the output hex file?
[04:46:13] <Igor2_> >grummund> you mean in textual form?
[04:46:42] <Igor2_> (my main problem was converting abstract things like clock source and freq to hex/bin and back all the time)
[04:47:46] <grummund> http://www.nongnu.org/avr-libc/user-manual/group__avr__fuse.html
[04:48:35] <Igor2_> nice
[04:48:41] <Igor2_> however, this is per-bit setting still
[04:48:55] <Igor2_> what i wanted to achieve is something really readable and setting based, not bit based
[04:49:09] <grummund> yes, i was just mentioning it for interest
[04:49:24] <Igor2_> thank you, i didn't know about this specific feature
[04:50:03] <grummund> RikusW is writing something similar
[04:51:22] <Igor2_> can you give me an url to that project?
[04:51:38] <Igor2_> unfortunately i need to leave now for a few hours, bit will leave my irc client here logging turned on,brb
[04:53:49] <grummund> http://sites.google.com/site/megau2s/home
[06:15:39] <norbi> hello
[06:16:10] <norbi> im getting these values from adxl345 http://pastebin.com/ymFr8gux
[06:16:26] <norbi> then divided them to 57 minimum scale for LSB to get g's
[06:16:36] <norbi> data format is on 8g
[06:18:37] <norbi> but im getting now weird values: http://pastebin.com/n2JEFDLw
[06:18:50] <norbi> shouldnt these values be max 1g?
[06:34:38] <mouche> I'm trying to work with the xplained xmega-a1 board and the drivers aren't working in windows 7. anybody know about this problem?
[07:01:57] <Kevin`> I remember hearing that the default software on the bridge chip was a bit wacky
[07:02:25] <mouche> Kevin`: that's encouraging
[07:02:43] <Kevin`> abc* would know, assuming it's not just a silly windows issue
[07:10:05] <mouche> abcminiuser: I hear you have experience with the xmega-a1 board
[07:11:58] <norbi> guys, how can i generate a register map?
[07:12:13] <norbi> i want to see each bit address of a memory
[07:13:13] <Kevin`> norbi: do you mean with jtag or equivalent, or from software? not really sure what you want
[07:14:24] <norbi> Kevin`: just in software
[07:14:39] <norbi> Kevin`: i want to see the addresses clearly
[07:15:12] <norbi> some sort of excel table, 8 colons(bits/colon) and rows as long as registers
[07:15:46] <abcminiuser> Hi mouche
[07:15:53] <abcminiuser> Indeed, several versions of it
[07:15:55] <norbi> im afraid there are some directly addressed bits in this code, because i cant find that register address
[07:16:09] <Kevin`> norbi: I think either most or all of the registers are mapped into memory, if you want to loop through them and print them out
[07:16:27] <norbi> no, i dont want to do that
[07:16:32] <norbi> i just need a map
[07:16:32] <abcminiuser> ^ Kevin` , that's abcminiuser BCS, BEE now :)
[07:16:34] <norbi> of the memory
[07:16:42] <Kevin`> norbi: oh. the datasheet has that
[07:16:49] <norbi> bit 0x000001, bit ..... etc
[07:16:56] <norbi> Kevin`: no it doesnt
[07:17:05] <Kevin`> abcminiuser: actually, ab* is a valid match too :D
[07:17:06] <norbi> Kevin`: im talking about adx345
[07:17:31] <abcminiuser> Kevin`, I suppose regexes are allowed in here then :P
[07:17:37] <abcminiuser> mouche, which board exactly do you have?
[07:17:44] <norbi> i need to see exaclty the bit adress not the bye
[07:18:09] <mouche> abcminiuser: Xplained XMEGA-A1 Rev 07
[07:18:23] <abcminiuser> Green PCB, or Blue?
[07:18:24] <Kevin`> norbi: what hardware addresses are there that point to bits?
[07:18:32] <mouche> Blue
[07:18:41] <abcminiuser> Ok, so the newer ones then
[07:18:45] <mouche> Yes, bought this week
[07:19:01] <abcminiuser> Okiedokie, well the default bridge should work with Windows, did you give it the supplied INF driver file?
[07:19:10] <norbi> 0xF2 for example
[07:19:21] <Kevin`> norbi: that's a value
[07:19:22] <mouche> Yes
[07:19:43] <norbi> where do i know from that it is a value or an address?
[07:20:10] <Kevin`> well, what I mean is it's a number, you haven't given enough information about it
[07:20:14] <abcminiuser> mouche, did the driver install?
[07:20:30] <norbi> this is the code: http://pastebin.com/93taNGtF
[07:20:33] <Kevin`> norbi: I assume you know how to do the equivalent of this with your windows calculator? http://pastebin.com/n42AfTbR
[07:20:38] <mouche> abcminiuser: it failed the first time
[07:20:41] <norbi> i dont really understand how does it reads out x, y,x
[07:20:46] <norbi> z*
[07:21:31] <norbi> kevin sure,
[07:21:31] <mouche> abcminiuser: I ended up finding this blog http://grieg.gotdns.com/blog/?p=545 and got the XML file so the XMEGA-A1 would show up in FLIP
[07:22:07] <norbi> but how do i know the difference between the address and value?
[07:22:27] <Kevin`> norbi: by how it's used. in sane programs you also know by what it's called.
[07:22:34] <Kevin`> norbi: what part are you confused with
[07:23:34] <abcminiuser> mouche, wha?
[07:23:40] <abcminiuser> The board has two controllers
[07:23:42] <norbi> mostly with this: http://pastebin.com/Pn9qXKML
[07:23:44] <abcminiuser> One, the XMEGA
[07:23:55] <abcminiuser> And two, an AVR32 for the USB passthrough
[07:23:59] <mouche> abcminiuser: right
[07:24:08] <abcminiuser> The driver should be for the AVR32, making it appear as a virtual COM port
[07:24:21] <Kevin`> norbi: what about it?
[07:24:23] <norbi> Kevin`: adxl is ss pin
[07:24:33] <norbi> i dont really understand what is he doing there
[07:24:34] <mouche> ok, so what's the next step?
[07:24:56] <Kevin`> norbi: it's just setting/clearing bit 0 in the register PORTB
[07:24:59] <norbi> i understand in most that he reads out the x,y,z registers, but where is the addres of it
[07:25:19] <Kevin`> norbi: I don't see x/y/z registers anywhere
[07:25:40] <Kevin`> and it's not like it's assembly where you have to do crazy stuff like that to make logic
[07:26:03] <norbi> they are stored in data[i]
[07:26:13] <abcminiuser> mouche, once you're up and running with the virtual COM port, you should be able to use FLIP from the command line to reprogram the XMEGA, if that's your thing
[07:26:16] <norbi> data[0],data[1] is x
[07:26:19] <norbi> next y
[07:26:22] <abcminiuser> Or you can reprogram it with an external programmer
[07:26:29] <Kevin`> norbi: don't call it a register. also what about it?
[07:26:59] <norbi> and so one, but how does he gets these? or he doesnt reeds the specific address where these values are stored? the slave automatically sends them?
[07:27:44] <Kevin`> norbi: it looks like it's getting the information in data[] from spi
[07:28:02] <mouche> abcminiuser: ok. so should I get the hex file by compiling files in AVR Studio 5 with the xmega128A1 simulator or is there a Makefile for it somewhere?
[07:28:08] <norbi> yea, but on spi, it should send first the reading address, right?
[07:28:17] <norbi> or the ss is sending it automatically?
[07:28:30] <norbi> i dont need to read it out from a registry?
[07:28:33] <Kevin`> norbi: depends on the device. it's sending "F2" first thing
[07:28:44] <Kevin`> norbi: spi doesn't define a command format like that
[07:28:54] <norbi> yes, but i definetly dont understand why
[07:29:00] <Kevin`> hopefully you have a datasheet for the thing it's connected to? you should read it
[07:29:10] <abcminiuser> mouche, compile with AS5, don't need the simulator
[07:29:10] <norbi> ive read it many times
[07:29:12] <Kevin`> *first though
[07:29:17] <norbi> but this code is just weird for me
[07:29:22] <norbi> its not mine
[07:29:30] <norbi> im just trying to analyze i
[07:29:30] <norbi> t
[07:29:31] <abcminiuser> AS5 comes with the AVR Toolchain, so if you build your project within the IDE, it'll make the HEX for you
[07:30:33] <norbi> values got from adxl345 accmeter are changeing just from 10 to 10 seconds, this is the really main problem, and im trying to debug it
[07:30:34] <mouche> abcminiuser: ok great. so this is programming the XMEGA A1 chip through the bridge chip transparently?
[07:30:44] <Kevin`> norbi: it's activating CS, sending 0xF2 to spi (and reading nothing), then reading 6 bytes into data[], then releasing cs. now what's confusing?
[07:32:01] <abcminiuser> mouche, not exactly
[07:32:23] <abcminiuser> Assuming the A1 was preloaded with a bootloader, you need to compile with AS5, then use FLIP from the command line to reprogram the XMEGA with the generated HEX file
[07:32:28] <Kevin`> norbi: .. and further, the device is sending back 3 16-bit... floats?, x, y, then z
[07:32:40] <Kevin`> maybe not floats
[07:32:53] <norbi> Kevin`: why it sends 0xF2, i cant find anything about that in datasheet, then why values in spi are the same for aprox 10 seconds
[07:33:12] <norbi> Kevin`: exactly
[07:33:28] <mouche> abcminiuser: ok. what protocol is it programming with?
[07:33:29] <abcminiuser> Whelp, I think I broke my site backend
[07:33:31] <Kevin`> norbi: I don't know why, that would depend on the device you are talking to.
[07:33:37] <abcminiuser> Or accidentally triggered a security system
[07:33:44] <mouche> heh, whoops
[07:34:02] <abcminiuser> mouche, FLIP uses DFU, over serial for the XMEGA-A1
[07:34:44] <abcminiuser> Damnit, last time I try to add another joke to my site
[07:34:51] <mouche> abcminiuser: ok, I'll look into that. Do I need to make sure I configure the UART port to make sure I don't remove the USB serial connection?
[07:34:54] <norbi> Kevin`: yea, could you please look at the adxl345 just a bit?
[07:35:21] <abcminiuser> mouche, no, the bootloader resides in a different section of the XMEGA
[07:35:27] <abcminiuser> (If one was loaded at the factory)
[07:35:40] <abcminiuser> You can't (easily, at least) use the bootloader to overrite itself
[07:35:53] <Kevin`> norbi: I did mention sbi and cbi are debrecated as c functions, right? using them with a recent compiler isn't harmful, although it's misleading
[07:36:50] <mouche> abcminiuser: but the AT32UC3B1256 allows the USB connection to work by creating a USB-to-UART bridge, right? That's what I've read. If so, don't you need that setup correctly for the programming to work?
[07:37:50] <abcminiuser> mouche, yes, the AVR32 performs the USB to USART conversion. When you want to reprogram you'd have to start the XMEGA's bootloader, which is a seperate program in the XMEGA and will reconfigure the USART as needed
[07:38:02] <abcminiuser> Your application doesn't have to make use of the USART at all, it's completely independant
[07:38:35] <mouche> oh I see. So from what I've read, you start it by holding SW0 before connecting via USB
[07:38:49] <mouche> then you can unplug and plug it back in without hitting the SW0 and it will run normally
[07:39:21] <norbi> Kevin`: sbi and cbi are redefined in my code as: #define sbi(var, mask) ((var) |= (uint8_t)(1 << mask))
[07:39:26] <abcminiuser> That would sound about right
[07:39:28] <mouche> so the first boot holding SW0 runs the bootloader which sets up the USART correctly and can program the chip, and then otherwise it runs the normal program
[07:39:36] <abcminiuser> Yes, exactly
[07:39:45] <mouche> thanks, that helps a lot
[07:39:49] <abcminiuser> The chip will have been preloaded with the bootloader program, which runs on startup
[07:40:01] <abcminiuser> That will check the button, and if not pressed with jump to the user application
[07:40:17] <abcminiuser> If pressed, it will stay in the bootloader and await programming instructions relayed through the AVR32
[07:40:40] <Kevin`> norbi: in addition to the address, there's a read/write bit and a multi-byte bit
[07:41:12] <mouche> ok, makes sense. so I built the sample program for the board and it gave me a .elf. Is that the right file to give to FLIP to program?
[07:42:14] <Kevin`> norbi: logically, from what it's doing, multi-byte works similarly to i2c where it increments the address for the other bytes (this matches the register map in the datasheet for x,y,z)
[07:42:48] <mouche> abcminiuser: it seems to be only looking for Intel HEX and AVR A90 files
[07:43:09] <norbi> Kevin`: i dont understand, this is a bit high for me
[07:43:14] <abcminiuser> mouche, AS5 should generate an ELF file (for debugging) and a HEX file (for programming)
[07:43:48] <Kevin`> what it's sending is this: not write + multibyte + register 0x32
[07:44:45] <norbi> where he does that?
[07:45:13] <Kevin`> 32 plus read plus multibyte is F2
[07:45:18] <Kevin`> try it in your calculator
[07:45:38] <norbi> where is in datasheet that part where it describes the read and multibyte?
[07:46:11] <Kevin`> norbi: where it's first mentioned, and also the timing diagram
[07:47:14] <mouche> abcminiuser: well it wasn't generating the .hex in Debug or Release mode, but I found an option to generate it in the Build settings
[07:49:04] <norbi> oh i understand now this schematic
[07:49:29] <abcminiuser> mouche, *sigh*, AS5 :P
[07:50:24] <mouche> abcminiuser: heh, well at least I got it working. whew, 50 seconds to program
[07:50:49] <norbi> it writes on 0 and reads on 1 on r/w bit
[07:50:59] <norbi> then need to set the mb bit
[07:51:15] <norbi> + 5bit address
[07:52:23] <norbi> so every time i need to read i need to specify first the address, then read, and same way in the write
[07:53:23] <Kevin`> yeah, that device apparently doesn't send anything meaningful while you are sending the address (obviously not the data, but sometimes they will give you status info right away)
[07:54:05] <mouche> abcminiuser: also, someone I talked to earlier mentioned that you know a bit about USB. are you familiar with USB HID?
[07:54:08] <norbi> Kevin`: ok, now i understand whats happening here
[07:54:20] <norbi> but one more thing is unclear for me
[07:54:27] <abcminiuser> mouche, somewhat: www.lufa-lib.org
[07:54:32] <norbi> why it passes 10 seconds before next refresh?
[07:55:39] <mouche> abcminiuser: what have you done with it?
[07:55:52] <Kevin`> norbi: maybe just because it's a demo. the device is able to send data relatively quickly
[07:56:10] <mouche> abcminiuser: oh, haha. you wrote it...
[07:56:14] <abcminiuser> mouche, just wrote a complete USB stack for the USB AVRs, including a HID driver and examples...
[07:59:10] <mouche> abcminiuser: wow. so I've never done USB before, and I'm interested in communicating with an Xbox 360 controller
[07:59:55] <mouche> I've looked at the linux driver for it, and they have the format of the data the controller sends via USB. It's an HID device. I've been looking for tutorials/help on where to go from there...
[08:00:43] <Kevin`> doesn't the xbox 360 controller appear as a STANDARD gamepad-type device, aside from the extra voice crap
[08:00:46] <Kevin`> ?
[08:01:39] <norbi> Kevin`: no more problem it seems
[08:02:34] <mouche> Kevin`: I think so. It falls under HID (Human Interface Devices), which includes gamepads.
[08:03:10] <Kevin`> mouche: some stuff uses HID without a defined interface, simply because it's one of the only things you can use without a "driver" on windows systems
[08:04:50] <mouche> well, I know the format that the data is sent in because it's available in the xboxdrv userspace driver. I'm not sure what all I need to get to the point where my code can just receive bytes and load in each message into a big struct
[08:04:58] <abcminiuser> mouche, haven't done that myself, but there is a LUFA project that emulates a XBOX 360 controller
[08:05:12] <mouche> I was considering using the vinculum II as a USB to USART translator, but I'm not sure how to figure out what to do next
[08:05:15] <abcminiuser> Note that even though it might use HID, it might not be as regular as you think
[08:05:38] <abcminiuser> For instance, the PS3 controller is standard HID, but requires some funky commands to be sent to it before it'll do anything
[08:05:42] <mouche> abcminiuser: ok, well I'm wanting to have the uC act as a host and get the controller's state, not act as a controller for the Xbox
[08:06:08] <abcminiuser> mouche, in that case, you need a USB AVR that supports USB Host mode
[08:06:31] <abcminiuser> Which means pretty much all of the AVR32s, plus the AVR8 devices of AT90USB1287 or AT90USB647
[08:07:40] <mitsakos> Hello! i'm trying to use uart on atmega8. I initilize it with F_CPU=8000000 (External Crystal High Freq fuse) and UBRR = 51 which is about 9600 baud rate. I try sending a char and i get it bat the baud rate it comes is 314! that's what my logic analyzer shows and i can't read this value from a pc serial port
[08:08:20] <Kevin`> mitsakos: F_CPU isn't a variable, it's a define that's used in some libraries
[08:08:22] <norbi> Kevin`: thank you1
[08:08:23] <norbi> !
[08:08:38] <mitsakos> yes i know it is defined in makefile
[08:09:02] <mitsakos> i also use a pwm output and it works fine..
[08:09:16] <Kevin`> norbi: put in a good word for me ;p
[08:28:41] <norbi> Kevin`: i will :D
[08:29:04] <norbi> z axxis on acc doesnt have that big range
[08:29:15] <norbi> i cant see it changing that fast like x,y
[08:29:26] <norbi> z behaves different
[08:48:15] <mitsakos> i'm using the 16-bit timer for fast pwm generation with ICR1 top value. Clock prescaler 8. 8000000/8 = 1Mhz. So the pwm output supposed to be 1mhz? The pwm output measured with oscilloscope is 4khz. It looks like the clock source is 32khz?
[08:48:29] <mitsakos> also ICR1 = 255
[08:49:39] <mitsakos> w8 i think i'm wrong
[08:49:40] <mitsakos> :P
[08:50:41] <Kevin`> mitsakos: the prescaler sets the clock for the timer, you are forgetting to divide by top-bottom
[08:51:09] <mitsakos> yes yes i found the formula
[08:51:37] <mitsakos> so it is correct. I have output of 3.9khz with prescaler 8 and external crystal at 8mhz
[08:51:54] <mitsakos> now i'm sure that the uC clock is correct and is at 8mhz
[08:52:41] <Kevin`> I don't suppose you have one of those nice chips that has a pll you can attach to the peripheral clock to run it faster than the rest of the system, no?
[08:52:50] <mitsakos> now about the uart. I use UBRR = 51 as the datasheet writes this is the value for 9600 baud rate. But what i get is 314
[08:53:13] <Kevin`> 51 decimal or 51 hex
[08:53:26] <mitsakos> i use 51 decimal
[08:54:00] <mitsakos> does it has to be hex?
[08:54:19] <Kevin`> I dunno, you are the one looking at the datasheet
[08:55:14] <mitsakos> it has a table and i has these numbers but there isn't any note explaining if it is decimal or hex
[08:56:15] <mitsakos> UBRRH = (uint8_t)(51>>8); // set baudrate counter
[08:56:15] <mitsakos> UBRRL = (uint8_t)(51);
[08:56:22] <mitsakos> this is how i enter the UBRR
[08:56:51] <Kevin`> 51 is less than 2^8
[08:57:09] <Kevin`> but anyway, I really won't go look at how UBRR is supposed to be set right now :)
[08:57:17] <Tom_itx> http://www.wormfood.net/avrbaudcalc.php
[08:58:32] <OndraSter> yo dawg, eya
[08:58:51] <Kevin`> xmega has a fractional baud generator, right?
[08:59:03] <mitsakos> yes i know i have used it in this way in order to replace 51 with a define value
[08:59:05] <Kevin`> it's kind of a shame there's no dip parts, there's so much new sanity that went into those
[09:01:09] <mitsakos> Tom_itx: so my UBRR is correct
[09:03:42] <Tom_itx> i don't think the mega8 has a ckdiv8 fuse does it/
[09:03:43] <Tom_itx> ?
[09:04:35] <Tom_itx> also, i don't know why people insist on still using that chip
[09:05:06] <Kevin`> because they have always used it
[09:06:06] <Tom_itx> it's obsolete
[09:06:11] <mitsakos> no it doesn't
[09:06:32] <Tom_itx> mega48 88 168 328 have superseeded it
[09:06:45] <mitsakos> what to use instead
[09:06:56] <Tom_itx> ^^
[09:07:05] <mitsakos> :P
[09:07:07] <Tom_itx> use it if you have it but move on
[09:08:05] <mitsakos> well my problem now is the uart :P
[09:08:27] <mitsakos> why it works in 314 baud rate?
[09:08:34] <OndraSter> what about attiny13a :P
[09:08:37] <OndraSter> that is superobsolete
[09:08:42] <OndraSter> so it is being sold out cheaper than cheap
[09:08:49] <OndraSter> sure, no SPI, no UART... :D
[09:08:52] <inflex> ??? 13A is not obsolete
[09:08:55] <inflex> it's quite new in fact
[09:08:58] <OndraSter> uh
[09:08:59] <inflex> replaces the T13
[09:09:01] <Tom_itx> mitsakos, http://tom-itx.dyndns.org:81/~webpage/how_to/atmega168/mega168_rs232_index.php
[09:09:04] <Tom_itx> inflex
[09:09:06] <Tom_itx> hello
[09:09:10] <inflex> hiya Tom_itx
[09:09:17] <Tom_itx> when does gp go on vacation?
[09:09:17] <OndraSter> huh
[09:09:21] <OndraSter> I must have checked attiny13
[09:09:24] <OndraSter> 's datasheet
[09:09:33] <inflex> Tom_itx: hmmm... good question that, not sure
[09:09:42] <Tom_itx> i need to send one in
[09:09:43] <inflex> OndraSter: anyhow, not that the 13A adds much, only a few power-register bits
[09:09:45] <OndraSter> but... no SPI, no UART
[09:09:47] <Tom_itx> but don't wanna get stuck
[09:09:53] <inflex> OndraSter: it's a Tiny...
[09:09:53] <OndraSter> SPI just for programming
[09:09:56] <OndraSter> well yes
[09:10:01] <OndraSter> but it is hard to connect it to outter world
[09:10:03] <inflex> 8 pins...
[09:10:08] <inflex> OndraSter: more than ample for a lot of things
[09:10:09] <OndraSter> and it has SPI for programming, so why not for general usage
[09:10:21] <OndraSter> sure, software SPI is not that hard
[09:10:27] <inflex> OndraSter: hope you don't ever look at the T10 ;)
[09:10:50] <Tom_itx> no kidding
[09:10:59] <OndraSter> 6pin o_O
[09:11:11] * inflex uses a lot of the T10's
[09:11:17] <inflex> and even more of the T13's
[09:11:34] <OndraSter> I thought about getting bunch of 13a
[09:11:40] <OndraSter> to make small measuring stations
[09:11:47] <OndraSter> with SPI that would send data through wireless chip
[09:11:55] <OndraSter> and this data would be obtained by 1wire
[09:12:00] <inflex> 13 or 13A, won't make much difference... T10's are great in that they have a nice 16-bit timer, but lack ADC resolution (8-bit)
[09:12:03] <OndraSter> from digital thermometer
[09:12:10] <Tom_itx> http://tom-itx.dyndns.org:81/~webpage/temp/tiny/tinyTPI1.jpg
[09:12:12] <Tom_itx> t10
[09:12:33] <OndraSter> how do you programm these? dW?
[09:12:38] <inflex> TPI
[09:12:38] <Tom_itx> isp
[09:12:39] <Tom_itx> tpi
[09:12:41] <Tom_itx> sry
[09:12:57] <Tom_itx> with my programmer :)
[09:13:27] <inflex> OndraSter: http://nqrc.com/images/PLD-BAC-55C.jpg
[09:13:39] <inflex> OndraSter: works great for jobs like that
[09:13:39] <OndraSter> yay
[09:14:05] <Kevin`> inflex: what's that do?
[09:14:05] <Tom_itx> smallest thing on there does the most work
[09:14:51] <OndraSter> TinyProgrammingInterface
[09:14:52] <OndraSter> ohh
[09:14:54] <inflex> Kevin`: it's a 2S lipo to 5.5V regulator with lost model alarm, low voltage alarm and voltage beacon/readout
[09:15:25] <inflex> Tom_itx: I want to get a batch of boards done for the BAC-55C units but I can't :( (not enough $)
[09:15:38] <inflex> Tom_itx: faster/cheaper to build the BAC55's too, since it's all 1-sided
[09:15:44] <Tom_itx> i've been in that spot for quite a while
[09:16:02] <OndraSter> TPI looks like... serial sync
[09:16:10] <OndraSter> half duplexed
[09:16:13] <inflex> Tom_itx: yeah, it's really annoying :( I'm thankful that ITead/seeed exist though, their $14 boards lets me at least keep proto'ing
[09:18:54] <OndraSter> I got either the thing or money back if it didn't arrive
[09:19:02] <OndraSter> or if the seller got deactivated, money through paypal
[09:21:18] <Tom_itx> !thislog
[09:21:18] <tobbor> This one: http://rueshouse.dyndns.org:82/~ircjunk/irclogs/html/%23avr-2011-12-11.html
[09:39:42] <Tom_itx> everybody fall asleep?
[09:44:42] <inflex> I certainly am
[09:49:23] <nn7> I have some simple code that I can't get to work properly. I have three potentiometers, rail to rail, wired into the first three analog inputs, I have the OCR?B pins wired to LEDs.
[09:50:19] <nn7> changing the pot should change the LED brightness. That does work, but when I turn ADC1 down to zero volts and ADC2 down to zero volts, turning ADC0 up also turns the other two LEDs up (but not all the way)
[09:50:25] <nn7> here's my code: http://pastebin.com/5WSABqaY
[09:50:47] <nn7> man, pastebin does some wacky stuff with the formatting
[09:52:08] <inflex> g'nite all
[09:52:15] <Tom_itx> gnite
[09:52:37] * mitsakos is away: auto-away
[09:53:44] <nn7> I can verify with an oscope that ADC1 and ADC0 are at 0 volts
[09:53:56] <nn7> that makes me think that something is wrong with my code
[09:56:19] <landonf> me reading the amforth page: FORTH! avr support! low-resources! … GPL. Oh :(
[09:56:56] <nn7> commenting out lines 47 through 50 causes the other two LEDs to stop responding to ADC0... so that's as expected
[09:56:57] <Tom_itx> nn7, this is an adc example that sends the value out the usart: http://tom-itx.dyndns.org:81/~webpage/how_to/atmega168/mega168_adc_index.php
[09:57:15] <Tom_itx> check the adc settings
[09:57:29] <nn7> why do you bit-shift 0?
[09:58:27] <Tom_itx> you don't need to
[09:58:29] <inflex> It's more as a tutorial guide thing
[09:58:36] <Tom_itx> it's there to account for all the bits
[09:58:44] <Tom_itx> for some silly noob
[09:58:50] <inflex> right, off to bed, sorry --- bbl ... well, 7~8hrs ;D
[09:59:07] <nn7> goodnight
[10:00:25] <nn7> hmm... looks like it's some capacitive problem
[10:00:41] <nn7> if I do a double ADC on the same channel, I get blinking LEDs
[10:00:50] <Tom_itx> what's aref wired to?
[10:01:10] <Tom_itx> the first sample is usually invalid
[10:01:11] <Tom_itx> anyway
[10:01:29] <nn7> aref has a cap to ground
[10:02:04] <nn7> avcc has an inductor to vcc and a cap to ground
[10:02:26] * mitsakos is back (gone 00:09:48)
[10:02:35] <nn7> as recommended in the datasheet
[10:03:16] <nn7> I've just never seen it this bad before
[10:03:35] <Tom_itx> is it breadboarded?
[10:03:52] <nn7> the blinking LEDs look like they galloping
[10:03:54] <nn7> yes
[10:04:04] <Tom_itx> they are noted for causing problems
[10:04:11] <Tom_itx> check all your wiring
[10:06:08] <OndraSter> nn7, isn't it because
[10:06:10] <OndraSter> if you are on 0V
[10:06:18] <OndraSter> the OCR registers are 0
[10:06:27] <nn7> ohhhh
[10:06:34] <OndraSter> and they get to the max value (0) immediatily?
[10:06:35] <OndraSter> :)
[10:06:48] <OndraSter> just guess
[10:06:51] <OndraSter> let me check the code
[10:07:29] <OndraSter> yap
[10:07:33] <OndraSter> could be the cause :P
[10:08:20] <birdpoop> Hello, im trying to connect my stk500 to my pc but windows does not recognize the hardware and lists it under other device as USB-Serial controller, what should i do?
[10:08:51] <OndraSter> birdpoop, that is rather normal, or isn't it? I don't have stk500 (let me check manual for it), but there is some usb<> serial converter (as ftdi ft232rl for example)
[10:09:06] <OndraSter> nn7, oh wait, you have different issue
[10:09:13] <OndraSter> didn't read to the end lol
[10:10:01] <nn7> well, there's also the problem of ADCH going over 250
[10:10:16] <nn7> so I did some conditioning of the ADCH value and I did a double ADC
[10:10:34] <nn7> doing the double ADC before setting OCR?B eliminated the issues
[10:10:47] <Tom_itx> is the adc 8bit?
[10:11:27] <nn7> I use ADLAR and take the top byte, so yes
[10:11:54] <Tom_itx> the top byte will be 0 until the bottom one is full right?
[10:12:15] <nn7> yes
[10:13:18] <nn7> weird, ok, it's a capacitance issue
[10:13:20] <nn7> thanks guys!
[10:18:20] <birdpoop> well im very new to this and just following basic tutorials to connect it, and i cannot connect to it using avr studio, and i think normally my device should be listed under com ports in my device manager?
[10:29:36] <birdpoop> windows finds no driver for the device..
[10:32:57] <Tom_itx> look for one in /programfile/atmel.....
[10:40:18] <birdpoop> brb reboot
[10:49:26] * mitsakos is away: auto-away
[10:55:35] <norbi> guys, im getting this compiling error http://pastebin.com/NDbv7Rug
[10:55:49] <norbi> im trying to use square() function from math.h
[10:57:52] <nn7> where's the code?
[10:58:28] <rue_bed> to square a number?
[10:58:31] <Tom_itx> Pastebin.com is under heavy load right now
[10:58:45] <rue_bed> I think you should make your own macro
[10:58:46] <nn7> norbi, first google hit says to add -lm to libs
[10:59:02] <rue_bed> #define sqr(x) ((x)*(x))
[11:00:33] <Tom_itx> rue_bed when he wakes up: http://tom-itx.dyndns.org:81/~webpage/rue/rue_cat.jpg
[11:00:57] <norbi> i have -lm
[11:01:35] <nn7> found another suggestion about rearranging the linking order
[11:01:48] <rue_bed> #define sqr(x) ((x)*(x))
[11:01:57] <nn7> rue_bed, oh sure, go the easy route
[11:02:34] <norbi> R = square ( AccelX_float^2 + AccelY_float^2 + AccelZ_float^2) what does ^2 means?
[11:02:37] <norbi> complement of 2?
[11:03:14] <nn7> http://en.wikipedia.org/wiki/2s_compliment
[11:03:17] <rue_bed> thats square root!!
[11:03:35] <rue_bed> its not working cause you asking fo rhte wrong thing
[11:03:41] <rue_bed> you want sqrt(----
[11:04:00] <rue_bed> yea, your doing that all wrong
[11:04:46] <rue_bed> R = sqrt( sqr(AccelX_float) + sqr(AccelY_float) + sqr(AccelZ_float))
[11:04:49] <rue_bed> ;
[11:04:55] <rue_bed> mmmkay?
[11:05:07] <nn7> ahh.. normalizing a vector
[11:05:39] <rue_bed> nobody wants to use the library I made cept zhanx...
[11:07:44] <rue_bed> .... did that help or did I just confuse everyone into silence?
[11:07:55] <norbi> rue_bed: ive copy pasted that from a documentation
[11:08:04] <rue_bed> well its not C
[11:08:33] <Tom_itx> forth?
[11:08:46] <norbi> nn7: can you explain please that therm of normalizing a vector? it writes exactly that therm in doc
[11:08:53] <rue_bed> it could be c++ if ^ has been overridden
[11:09:04] <norbi> it is VB
[11:09:41] <rue_bed> norbi, its the length of a line drawn from 0,0,0 to that co-ordinate
[11:10:07] <nn7> when you normalize a vector, you're reducing it to unit length
[11:10:08] * rue_bed wonders if 4 dimentional lines have volume
[11:10:40] <Tom_itx> they sneak up behind you
[11:11:20] <nn7> norbi, http://www.fundza.com/vectors/normalize/index.html
[11:11:33] <rue_bed> I think 4d objects can change their volume without changing their shape
[11:12:05] <rue_bed> funny, sounds a bit like electrons behave
[11:13:56] <norbi> ok, thats nice, but this documentation describes a bit differently
[11:14:31] <norbi> and im getting the same effect http://pastebin.com/1YyqbArJ
[11:14:39] <rue_bed> it sounds like it puts it in terms you dont understand
[11:14:50] <rue_bed> so your porting vb code to an avr?
[11:14:54] <rue_bed> what are you up to?
[11:15:23] <nn7> norbi, yeah, you're determining the length and assigning that value to R
[11:15:35] <nn7> then you're dividing each component by the length
[11:15:47] <nn7> you'll find that accely+accelx+accelz=1
[11:16:07] <norbi> rue_bed: im not, ive just finaly found a small application document of this adxl345
[11:16:09] <rue_bed> 1 l
[11:16:52] <norbi> nn7: yea, but this method is the same as in the fundza.com's doc?
[11:16:59] <norbi> i mean it does the same?
[11:17:20] <nn7> it *is* the same
[11:17:36] <nn7> what you call "r" they call |a|
[11:17:48] <nn7> you have x = ax/|a|
[11:17:53] <nn7> you have ax = ax/r
[11:18:13] <nn7> sorry, they have x = ax / |a|
[11:18:17] <nn7> you have ax = ax / r
[11:19:22] <norbi> but they are calculating sqrt(square+square+square), im calculating square(^2+^2+^2)?
[11:19:38] <OndraSter> how can ax = ax/r? That is 1 = 1/r
[11:19:44] <OndraSter> (missed the discussion)
[11:20:13] <OndraSter> oh gotcha
[11:20:17] <nn7> the character 1 does look like an axe
[11:20:18] <OndraSter> ax /= r
[11:20:26] <norbi> OndraSter: ax is changeing value to ax/r
[11:20:36] <OndraSter> norbi, yeah, I was on math view rather programming view
[11:20:40] <OndraSter> sorry :D
[11:20:43] <nn7> :)
[11:20:55] <nn7> for all values of r equal to 1...
[11:21:12] <norbi> OndraSter: yea :)
[11:21:20] <norbi> but thats better
[11:21:35] <norbi> if you are on math view, on programming view is math view too :)
[11:21:38] <nn7> 1 + 1 = 3 for extremely large values of 1
[11:21:48] <rue_bed> :) I like that one
[11:21:59] <norbi> =))
[11:22:25] <norbi> question, why is that good for me if i normalize the vector?
[11:23:00] <rue_bed> becuase it lets you know how far away the end-effector is so you can set the elbow to the right able to reach it
[11:23:26] * mitsakos is back (gone 00:33:59)
[11:23:37] * nn7 blinks
[11:23:37] <rue_bed> or it can tell you how fast your going
[11:24:04] <rue_bed> or it can tell you the distance between the corners of a cube
[11:24:16] <rue_bed> or even a square
[11:24:31] <nn7> norbi, it gives you direction without distance
[11:24:34] * mitsakos is away: I'm busy
[11:24:45] <rue_bed> oh, sorry, I didn't notice the topic change
[11:25:21] <rue_bed> when did you change to direction?
[11:27:31] <norbi> hmm, very interesting, im geting lost
[11:27:40] <nn7> norbi, if you show us what you're trying to do with that accelerometer, we can probably tell you why the code is doing it
[11:27:46] <norbi> ive found now the angles
[11:27:48] <rue_bed> you got here lost
[11:28:04] <norbi> http://www.starlino.com/imu_guide.html
[11:28:07] <OndraSter> so guys, I am making project with atmega128a... but the 64kB memories are non-existant (at least on farnell), only 128kB... how come that it is cheaper to get 128kB*8 rather 64kB*8 :P
[11:28:51] <norbi> nn7: im trying to get angles and measure distance and speed
[11:29:15] <norbi> i want to know in what direction does the robot moves, what disntace, and what speed
[11:29:44] <nn7> from that link, "This triplet is often called Direction Cosine , and it basically represents the unit vector (vector with length 1) that has same direction as our R vector."
[11:30:20] <nn7> so you're calculating the cosine of the angle
[11:31:40] <keenerd> I need to try rewriting an app to use rational trig. Seems like a clever idea.
[11:31:50] <rue_bed> is it a 6 axis accel?
[11:32:19] <rue_bed> ah, I suppose if its not in someones hand you only need 3
[11:36:46] <norbi> 3 axis
[11:37:00] <norbi> (.text.avr-libc.fplib+0x24): relocation truncated to fit: R_AVR_13_PCREL against symbol `__subsf3' defined in .text section in c:/winavr-20100110/bin/../lib/gcc/avr/4.3.3/avr5\libgcc.a(_addsub_sf.o)
[11:37:10] <norbi> i tried to find out what this could be
[11:37:18] <norbi> it says add -lm to linker option
[11:37:33] <norbi> tryied in project configuration linker option add -lm
[11:37:38] <norbi> but that is already there
[11:43:06] <rue_house> hmm
[11:46:02] <norbi> i cant figure out htis error
[11:46:11] <norbi> it says libma.o
[11:46:35] <norbi> but if i add it manualy it will give me another error that im not really allowed to do that
[11:50:47] <nn7> does this have something to do with not linking the floating point version?
[11:51:46] <nn7> I really dislike avr studio 5
[11:52:06] <nn7> norbi, I think you want libm.a, not libma.o
[11:53:07] <norbi> yea
[11:53:21] <norbi> im using avrstudio 4
[11:53:33] <norbi> but i think the command line is the problem
[11:53:41] <norbi> avr-gcc -mmcu=atmega64 -Wl,-u,vfprintf -lprintf_flt -lm -Wl,-Map=Cerebot2.map Cerebot2.o -o Cerebot2.elf
[11:53:49] <norbi> -lm is not at the end
[11:53:55] <norbi> how can i put it to the end
[11:53:56] <norbi> ?
[11:54:05] <norbi> i mean in avrstudio
[11:58:18] <norbi> any idea? i have -lm in the cmd but still wont include libm.a
[12:03:09] <norbi> ok, figured out
[12:03:11] <norbi> it works now
[12:03:22] <norbi> but arcos() shoudl give me this low values?
[12:03:25] <norbi> 2.0000
[12:03:27] <norbi> 1.2222
[12:03:29] <norbi> ant so ?
[12:04:06] <rue_house> norbi, how about you put physics aside for a few mins and work on getting avr math working?
[12:05:39] <norbi> rue_house: it works
[12:05:47] <norbi> rue_house: math works now
[12:06:32] <norbi> but i want to find out angle in degrees
[12:07:09] <rue_house> my lib includes two macros
[12:07:15] <rue_house> rads2degs
[12:07:19] <rue_house> and degs2rads
[12:07:32] <rue_house> #define rad2deg(K) ((K)*(180.0/3.1415926535))
[12:07:32] <rue_house> #define deg2rad(K) ((K)*(3.1415926535/180.0))
[12:08:07] <rue_house> see how to use them?
[12:09:08] <nn7> norbi, what did you do to fix it?
[12:10:03] <rue_house> wold anyone like to write a counterfunction for me?
[12:10:10] <rue_house> #define xy2i(X,Y,I,J) (((I)*(Y))+(X))
[12:10:17] <norbi> nn7: added libm.a mannually
[12:10:20] <rue_house> I'd like i2xy(
[12:10:24] <OndraSter> is the pi correct?
[12:10:32] <OndraSter> ok it is :)
[12:10:44] <norbi> rue_house: it looks like its almost there
[12:11:07] <norbi> but, degrees are relative to what?
[12:11:14] <norbi> to R?
[12:11:30] <rue_house> 3.141592653589793238462643383279502884197169399375105820974944
[12:11:48] <rue_house> all math fn's use rads
[12:12:00] <rue_house> so convert back and forth to degs as you like
[12:17:37] <norbi> ok, but the values i get are not correct, as i can see
[12:19:02] <norbi> this is in plain mode, not rotated: http://pastebin.com/SUh29pPD
[12:19:47] <rue_house> qI have a point manipulation library, you want it?
[12:20:09] <rue_house> 3.14 isn't accurate enough
[12:20:25] <rue_house> esp cause you didnt' make it handle the integers right
[12:20:37] <norbi> rue_house: ?
[12:21:13] <rue_house> 180/3.14 is 57 if you do it wrong
[12:21:27] <rue_house> or, depending, 60
[12:23:07] <rue_house> use the macro like I gave it to ya
[12:23:19] <norbi> im using it noe
[12:23:20] <norbi> now
[12:23:24] <norbi> it looks correct
[12:23:38] <norbi> bu i would like that in reference to z
[12:23:46] <rue_house> 180 is an integer, 180.0 is a floating point number
[12:23:47] <norbi> it not to be 90 degreees
[12:24:01] <norbi> i want to be z the 0 or 180
[12:24:58] <norbi> now if i let it horizontaly, it shows 90, if i move x,y direction to top, it is near to 0, if to bottom it is near 180
[12:35:44] <norbi> it looks like it works fine:)
[12:36:03] <norbi> now i just have to transform into 360 degrees
[12:40:01] <norbi> why is this happenning with y axis? http://pastebin.com/TZzD9AET
[12:40:10] <norbi> it looks like x axis degrees are smooth
[12:40:14] <norbi> but y is not
[12:40:17] <norbi> why is that?
[12:40:28] <norbi> it has some fluctuations
[12:40:50] <landonf> heh, do the contiki guys know that their website is totally (literally!) unnavigable
[12:40:53] <landonf> +?
[12:41:03] <landonf> http://www.contiki-os.org/
[12:41:33] <norbi> and if turn it around, then x will have those fluctuations
[12:45:21] <Tom_itx> norbi, adc?
[12:45:38] <Tom_itx> maybe you are getting bad readings or noise?
[12:45:53] <Tom_itx> or is it an spi sensor?
[12:46:10] <Tom_itx> spi you need to flush the last byte out
[12:47:26] <norbi> Tom_itx: i dont really understand what that means?
[12:47:47] <Tom_itx> what sensor is it?
[12:47:49] <norbi> it happens just when i turn it vertycally
[12:47:54] <norbi> just on vertycal position
[12:47:58] <norbi> adxl345
[12:48:04] <Tom_itx> is it spi?
[12:48:14] <norbi> yest it is 4 wire
[12:48:29] <Tom_itx> are you flushing the last byte out?
[12:48:31] <Tom_itx> of the slave
[12:48:35] <norbi> no
[12:48:47] <norbi> how should i do that?
[12:48:50] <Tom_itx> and discarding the first one
[12:48:58] <Tom_itx> send a dummy byte
[12:49:10] <Tom_itx> it may not apply to that sensor but generally speaking...
[12:49:18] <Tom_itx> that's the nature of spi i believe
[12:49:48] <Tom_itx> i haven't looked at that sensor
[12:51:12] <norbi> Tom_itx: here is the code, and send_dummy function
[12:51:13] <norbi> http://pastebin.com/8gbZsNZF
[12:51:18] <norbi> where should i use it?
[12:51:27] <norbi> the send_dummy how should be used here?
[12:53:48] <Tom_itx> why do you send 0?
[12:53:54] <norbi> where?
[12:53:57] <norbi> ah
[12:53:59] <Tom_itx> at the top
[12:54:16] <Tom_itx> typically you send 0xFF unless you are initializing the slave or something
[12:54:35] <norbi> ok, i will change it to ff
[12:54:40] <Tom_itx> probably doesn't matter if the slave doesn't expect data
[12:55:05] <Tom_itx> cpol etc set right?
[12:55:18] <Tom_itx> it's been a while since i've done spi..
[12:55:28] <norbi> yes, it is set right
[12:55:32] <norbi> cpol 1
[12:55:35] <norbi> cpah 1
[12:56:16] <norbi> Tom_itx: ive seen that most people novadays are using i2c instead, but i dont understand why...
[12:56:45] <norbi> Tom_itx: anyway, how should i put there send_dummy?
[12:57:06] <Tom_itx> is the 'for' loop where you get the data?
[12:57:11] <norbi> yes
[12:57:30] <Tom_itx> send an extra byte after it
[12:57:41] <Tom_itx> what's spi(PORTB... for?
[12:57:45] <Tom_itx> sbi
[12:57:51] <norbi> SS
[12:57:54] <norbi> CS
[12:57:55] <Tom_itx> ok
[12:57:55] <karlp> chip select
[12:58:30] <Tom_itx> low active slave?
[12:58:57] <norbi> this way? http://pastebin.com/cK6qYRg6
[12:58:59] <norbi> yes low active
[12:59:58] <Tom_itx> i'd put it right after the for }
[13:00:49] <Tom_itx> i need to look at my old spi stuff
[13:02:06] <Tom_itx> is data[0] good data?
[13:02:38] <norbi> this way? http://pastebin.com/pw3xWDHt
[13:02:51] <norbi> it looks like it works now, this way
[13:03:05] <Tom_itx> is that better?
[13:03:10] <norbi> no im wrong
[13:03:13] <norbi> it was at begining
[13:03:17] <norbi> but now the same
[13:03:42] <Tom_itx> check the first and last byte send and see
[13:03:47] <Tom_itx> one may not be good data
[13:03:59] <norbi> http://pastebin.com/Lddrrh1g
[13:04:10] <norbi> now like x and y constantly changes places
[13:04:12] <Tom_itx> you know i can't see pastebin on this pc...
[13:04:29] <norbi> hmm
[13:04:30] <Tom_itx> it crashes my brouser for some damn reason
[13:04:38] <norbi> any other pastebin?
[13:04:40] <norbi> codepad?
[13:05:17] <Tom_itx> without testing i'm not sure but i'd play around with sending an extra byte
[13:05:23] <Tom_itx> either at the beginning or end
[13:05:31] <Tom_itx> and maybe capture an extra byte in your array
[13:05:40] <Tom_itx> to see which one is bad
[13:06:08] <Tom_itx> i'm a bit rusty on spi
[13:07:05] <Tom_itx> is the 0xF2 an init sequence for the slave?
[13:07:06] <Tom_itx> at the top
[13:07:19] <norbi> http://codepad.org/2BGV7CZf
[13:07:44] <norbi> http://codepad.org/ImGyfF1i
[13:08:16] <norbi> Tom_itx: no, that is the R/W bit, and MB, and address
[13:08:33] <Tom_itx> you don't need to send a dummy both places
[13:08:40] <Tom_itx> i don't think
[13:09:08] <Tom_itx> if you do, you will loose data on one end or the other
[13:09:18] <norbi> do i need sbi and cbi before and after send_dumy
[13:09:20] <norbi> ?
[13:09:26] <Tom_itx> no
[13:09:30] <Tom_itx> leave it active
[13:09:35] <Tom_itx> i think
[13:11:51] <Tom_itx> the first xA looks like garbage
[13:12:34] <norbi> now im getting this: http://codepad.org/382070YI
[13:12:56] <norbi> with this:: http://codepad.org/DyuzviLM
[13:12:58] <Tom_itx> nope that's no good
[13:14:09] <Tom_itx> gotta run...
[13:14:27] <Tom_itx> try putting it right before the SBI line outside the loop
[13:14:37] <Tom_itx> although that may not help either
[13:15:00] <Tom_itx> aarg i need to brush up on spi again
[13:19:31] <norbi> Tom_itx: you are leaving?
[13:19:37] <norbi> and it still wont work
[13:19:49] <norbi> i dont understand why are these fluctuations apear
[13:51:29] <norbi> im trying to get device id
[13:54:50] <norbi> im getting 40
[13:55:06] <norbi> instead of printf("\n");
[13:55:09] <norbi> aah
[13:55:13] <norbi> instead of E5
[14:01:22] <norbi> sry
[14:01:26] <norbi> my net interrupted
[14:01:33] <norbi> what did you say?
[14:03:51] <OndraSter> nobody said anything
[14:05:02] <norbi> ah ok
[14:05:03] <norbi> sry
[14:05:17] <norbi> any idea why id is returned as 40?
[14:05:30] <norbi> i probably not reading it as it needed
[14:07:04] <norbi> sbi(SS) , send_byte(0x00); send_byte(0xFF); data=spi; cbi(SS); printf("%x",data);
[14:07:09] <norbi> and it gives me 40
[21:39:39] <abcminiuser> Aww
[21:39:49] <abcminiuser> My server security is too high
[21:40:12] <ziph> Oh?
[21:40:51] <abcminiuser> I'm trying to add a joke SQL query to one of my site pages, but it's tripping some internal security module
[21:41:12] <abcminiuser> I wanted to make it redirect to something funny if the user tries to modify the fake string
[21:41:19] <inflex> lo there abcminiuser and ziph
[21:41:31] <abcminiuser> But apparently any valid SQL statement in a URL query causes it to return a FORBIDDEN HTTP code
[21:41:31] <ziph> inflex: Lo, how's things with you?
[21:41:31] <inflex> abcminiuser: almost ironic :p
[21:41:32] <Tom_itx> farside awakes
[21:41:40] <inflex> Johnny drop tables?
[21:41:45] <abcminiuser> Yahoyoy folks
[21:41:54] <inflex> it's bloody hot... and I have 66c in the bank... :\
[21:41:54] <abcminiuser> Something like that, I just like little jokes
[21:41:55] <ziph> abcminiuser: That's a bit sad.
[21:42:13] <ziph> abcminiuser: What if you were doing a codebin style site and someone tried to paste SQL code to show someone?
[21:42:43] <abcminiuser> Oh in the body is fine, I mean this: http://fourwalledcubicle.com/Contact.php?sql=SELECT+*+FROM+Test
[21:43:16] <abcminiuser> My .htaccess hasn't got anything unusual in it to do that, so I've no idea where the logic is for it
[21:43:26] <abcminiuser> But I suppose I can't complain about too much security...
[21:44:27] <ziph> It's still a silly feature.
[21:45:07] <ziph> For starters the parser is probably naive enough for there to be some trivial work around.
[21:45:24] <abcminiuser> Indeed
[21:45:41] * inflex wonders if there's enough scrap change around the house to get something for dinner tonight
[21:46:22] <ziph> Did the cat just run? :)
[21:49:20] <inflex> hah
[21:50:01] <Casper> in a push pull forward, if the transfo is saturating, should I increase or decrease the turn count?
[21:52:03] <ziph> Saturation happens when you exceed a certain flux level, and reducing the turn could will reduce the flux. Is that what you mean?
[21:52:18] <ziph> reducing the turn count will reduce, rather.
[21:53:35] <ziph> It's also dependent on the signal you're putting in to it though.
[21:54:38] <Casper> I'm making a push pull forward, and I get a suddent drop in efficiency and heating of the transformer, increasing the input voltage barelly do anything
[23:01:48] <doublebeta> til nintendo is going to use avrs in their next gaming console