Firmware for the Spectral Multiband Resonator, a Eurorack-format module from 4ms Company.

The SMR is a 6-channel highly resonant filter ("resonator") with a ring of RGB LEDs, six sliders, and stereo audio inputs and outputs. The main processor is the STM32F427 chip. For more details see http://www.4mspedals.com/smr.php

The SMR would make a nice platform for other audio projects, the hardware contains:

• 180MHz 32-bit ARM chip with FPU (STM32F427)
• Stereo codec chip (running at 96kHz/16bit in this firmware)
  • Two audio inputs, two audio outputs
• Five 16-channel PCA9685 LED drivers (10-bits per channel)
• 26 RGB LEDs, 20 are arranged in a circle/ellipse, 6 are arranged in a line
• Six slider potentiometers with an LED on the shaft
• Five potentiometers
• Six momentary LED buttons
• A rotary encoder and button
• Six toggle switches (one is SP3T, others are SPST)
• Twelve CV inputs
• Six gate/digital inputs (can be used as unscaled CV inputs)
• Six CV outputs (0V-8V), driven by PWM

As of writing this (Sept, 2015), there are two PCB versions: 1.0 and 1.0.1. These are functionally identical and this firmware will run exactly the same on both versions.

You need to install the GCC ARM toolchain. This project is known to compile with arm-none-eabi-gcc version 4.8.3, and version 4.9.3.

It's recommended (but not necessary) to install ST-UTIL/stlink. Without it, you will have to update using the audio bootloader, which is very slow (5 minutes per update). With ST-UTIL or stlink and a programmer, you can update in 5-20 seconds. The Texane stlink package contains a gdb debugger, which works with ST-LINK programmers such as the STM32 Discovery boards to connect to the Spectral's 4-pin SWD header. The STM32F4 Discovery Board is low-cost (under US$15) and works great as a programmer and debugger.

You also may wish to install and IDE such as Eclipse. There are many resources online for setting up GCC ARM with Eclipse (as well as commerical software). This is totally optional. Instead of an IDE you can use your favorite text editor and a few commands on the command line (Terminal) which are given below.

Continue below for Max OSX, Linux, or Windows instructions.

For Mac OSX, follow these instructions to install brew and then the arm toolchain and st-link (taken from https://github.com/nitsky/homebrew-stm32):

ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)"
brew tap nitsky/stm32
brew install arm-none-eabi-gcc
brew install stlink

That's it! Continue below to Clone the Projects

For linux, check your package manager to see if there is a package for arm-none-eabi-gcc or gcc-arm-embedded or gcc-arm-none-eabi. Or just download it here:

Download GCC ARM toolchain from Launchpad

Next, install st-link from texane:

sudo apt-get install git libusb-1.0.0-dev pkg-config autotools-dev
cd (your work directory)
git clone https://github.com/texane/stlink.git
cd stlink
./autogen.sh
./configure
make
export PATH=$PATH:(your work directory)/stlink

The last command makes sure that the binary st-flash is in your PATH so that the Makefile can run it to program your module.

That's it! Continue below to Clone the Projects

Download GCC ARM toolchain from Launchpad

Download ST-UTIL

Install both. Please contact me if you run into problems that you can't google your way out of!

Make sure git is installed on your system. (OSX: type "brew install git" into the Terminal)

Create a work directory, and enter it.

Clone this project (SMR), stmlib, and the stm-audio-bootloader projects:

git clone https://github.com/4ms/SMR.git  
git clone https://github.com/4ms/stmlib.git
git clone https://github.com/4ms/stm-audio-bootloader.git

Create a symlink for stm-audio-bootloader so that it works with python (required to generate the .wav file for audio bootloading)

 ln -s stm-audio-bootloader stm_audio_bootloader

Verify that your directories are as follows:

(work directory)
|  
|--SMR/  
|--stm-audio-bootloader/  
|--stm_audio_bootloader/   <----symlink to stm-audio-bootloader
|--stmlib/

Make your changes to the code in the SMR directory. When ready to compile, make the project like this:

cd SMR
make

If you want to flash the file onto the SMR, attach an ST-LINK programmer in SWD mode using a 4-conductor cable and run:

make flash

Read the Discovery Board's manual for how to set it to program external devices. On the STM32F4 Discovery board, you have to remove two DISCOVERY jumpers to put it in ST-LINK mode. Connect the 4-pin cable to the SWD header starting at the dot (pin 1) on the Disco board. On the SMR, pin 1 is labeled with a "1". Any 4-pin cable or even 4 breadboard jumper wires will work. A 10-to-16 pin eurorack power cable is fine too, if it's short enough.

When ready to build an audio file for the bootloader, make it like this:

make wav

This requires python to run. It creates the file main.wav in the build/ directory. Play the file from a computer or device into the SMR by following the instructions in the User Manual on the 4ms SMR page.

Troubleshooting: If you have trouble getting python to create a wav file, such as this error:

ImportError: No module named stm_audio_bootloader

Then try this command:

export PYTHONPATH=$PYTHONPATH:'.'

The bootloader is a separate project, slightly modifed from the stm-audio-bootloader from pichenettes.

The bootloader is already installed on all factory-built SMRs.