Firmware hacking
Tides' source code is available under the MIT licence.
The code (along with the hardware description files) can be found in the tides2 directory in our Eurorack modules git repository.
After having cloned the repository, don't forget to run git submodule init && git submodule update to make sure the sub-projects referenced in the code are also pulled.
Toolchain
If you don't mind installing Vagrant and VirtualBox a cozy environment for firmware hacking is available.
If you want to set up your own environment to build Tides' code, an ARM EABI toolchain must be installed. Because of tight CPU and code size limits, we recommend you to use the same compiler version as we do: 4.8-2013-q4-major. Various pre-compiled binaries and source packages are available here.
The path to the toolchain binaries must be specified in the TOOLCHAIN_PATH variable in tides2/makefile.
To build the bootloader, use the following command:
make -f tides2/bootloader/makefile hex
To build the code, use the following command:
make -f tides2/makefile
If you modify lookup tables and want the big resources.cc file to be regenerated:
touch tides2/resources/resources.py && make -f tides2/makefile resources
Firmware programming
The recommended solution for firmware programming is to use a ST Discovery board. Connect the 4 SWD lines of the board to the 4 programming lines of Tides (RESET, SWDIO, SWCLK and GND).
Make sure that the programming interface (PGM_INTERFACE variable) is set to stlink-v2 in stmlib/markefile.inc.
Upload the firmware with:
make -f tides2/makefile upload
Note that this will destroy the factory calibration data.
DAC Calibration
With a factory-made module, it is recommended to backup and restore the data originally written in the Flash memory of the module in range 0x08004000:0x08008000 (factory calibration data).
In a DIY situation, the most direct way of calibrating the module is to initialize the structure storing the calibration data to the correct values, deduced by measurement. Modify Settings::Init() to include the following lines:
persistent_data_.dac_calibration[0].offset = 32768.0f /* FIXME */;
persistent_data_.dac_calibration[0].scale = -4032.9f /* FIXME */;
// SNIP SNIP
persistent_data_.dac_calibration[3].offset = 32768.0f /* FIXME */;
persistent_data_.dac_calibration[3].scale = -4032.9f /* FIXME */;