Modify the bootloader, or use an ISP programmer for all future programming.

where you write

Code:

if(memcmp(master, target_tag[i], TAG_LEN) == 0 )

is PROBABLY wrong

try

Code:

if(memcmp(master, target_tag, TAG_LEN) == 0 )

http://www.nongnu.org/avr-libc/user-manual/inline_asm.html

or you can just start writing everything in .S files and use the assembler through command-line

Regarding http://arduino.cc/en/Main/ArduinoWiFiShield

I'm mainly looking for the original firmware source code

I believe the firmware is found here

http://www.hd-wireless.se/index.php?option=com_content&view=article&id=46&Itemid=53

I'm not even sure if it is the same firmware as the one that is used in the Arduino Wifi Shield, but it must be pretty similar, similar enough for me to get some work done. Once I have that, it should be straight forward to compile it and flash the AVR32 using the ICSP.

But here's the problem, that website won't let you download anything without logging in. To create an account, their company has to manually approve each account. It's been about a week and I haven't received any replies. There are no accounts made available on bugmenot.com either.

I would appreciate it if anybody can provide the firmware source code.

Also I can't seem to find the instructions for using the DFU mode, but I think I'll be able to use ICSP just fine. I think somebody should add a page on this website explaining the procedure to help people out.

we're gonna need a circuit diagram, a text description might not be enough.

I might be misinterpreting what you are saying, but I have a theory anyways: power could be going through the TX line, just enough to power up the ATmega alone, but once it's in your circuit board, there's too many other components also drawing the same power and the ATmega does not have enough anymore to actually run code.

Code:

 Wire.beginTransmission(ADDRESS);
  Wire.write((uint8_t)0x00);
  Wire.write((uint8_t)0x00);
  Wire.endTransmission();

is wrong, what that does is "set pointer to 0x00, write 0x00 into where you are pointing"

each write auto-increments the pointer, that's why you are missing the first byte, because the write operation skipped it

Code:

 Wire.beginTransmission(ADDRESS);
  Wire.write((uint8_t)0x00);
  Wire.endTransmission();

is correct, what that does is "set pointer to 0x00, wait for a read operation from where you are pointing"

thanks everybody, I'm guessing it's the Bluetooth then

eried, that's quite an excellent mod

wow what happened to verbose? I didn't realize it disappeared until I saw your mod

It takes 20 seconds on average to get Arduino 1.0.1 to start. This is almost unacceptable. Does anybody else has this issue?

Before you blame my computer, it's a new laptop is running Win 7 Ultimate 64-bit, Sandy Bridge i7, 8 GB of DDR3 RAM, and a OCZ Agility 3 SSD that's on a SATA III port. SolidWorks starts faster than Arduino.

My latest purchase is a 20-foot shipping container.

What has this to do with Arduino/electronics?

Well I plan to fit half of it out as a workshop so when we aren't travelling I have somewhere decent to make stuff. At least I'll have a workbench now.

______
Rob

Mobile hackerspace? I'd support something like that.

@frank26080115 A separate DAC gives me the ability to create and send waveforms directly from the MCU as well. If I decide to implement something like a low frequency arb. wave generator then it will be easier to send the data directly to the DAC rather than packing it up into a format that the decoder will understand just for it to be decoded again.

VLSI Solution brand MP3 decoders can also play WAV files, and WAV files are just a series of amplitudes. It'll work kind of like a normal DAC when you tell it to play a WAV file.

edit: but it won't work quite like an arbitrary function generator though, so your point is still valid

I wanted to keep costs down and board layout simple. Seeing as it doesn't have to be realtime, software seemed like a fine solution

How are you planning to output audio?

http://www.analog.com/static/imported-files/data_sheets/AD5663.pdf
I have another topic in the audio section talking about it, as I'm not sure which SPI mode to use to talk to it.

You know, the MP3 decoder chips output audio directly, so you don't need the AD5663 anymore. The board layout is still very simple.

maybe libmad could be your starting point, i've seen it being used on a mbed, so it's not impossible to use

http://mbed.org/users/Gruenfrosch/code/madplayer/summary

it's declared in a class only if the library uses a class. if it is declared in a class, it must be public for you to access it. For example

If this is in the library's header file

Code:

class HardwareSerial
{
  public:
    int theVariableYouWantToAccess;
  private:
    int someOtherStuffYouCannotAccess;
};

extern HardwareSerial Serial;

then you can do something like

Code:

Serial.theVariableYouWantToAccess = 1;

inside your own sketch

if it's not inside a class, then you should be able to access it by using "extern" inside your sketch. For example

if your library has a .c or .cpp file that looks like

Code:

int theVariableYouWantToAccess;

void foo()
{
  theVariableYouWantToAccess++;
}

then your sketch can look like

Code:

extern int theVariableYouWantToAccess;

void setup()
{
}

void loop()
{
  theVariableYouWantToAccess = 0;
}

or you can have a header file that contains

Code:

#ifndef headerFile_h
#define headerFile_h

extern int theVariableYouWantToAccess;

#endif

and your sketch looks like

Code:

#include <headerFile.h>

void setup()
{
}

void loop()
{
  theVariableYouWantToAccess = 0;
}

this is an extremely unlikely reason, but since Arduino ships with an older and custom compiled version of WinAVR, maybe it's too outdated for this chip?

I expected the BOOTSZ options to be 'greyed out' when BOOTRST is disabled; but it wasn't.

There is a good reason for that: Because everything is extracted from the XML, and nothing in the XML makes a relationship between BOOTSZ and BOOTRST. The code isn't as smart as a human who would make any relationship between the two.