Skip to content

esphome/voice-kit-xmos-firmware

Repository files navigation

voice-kit-xmos-firmware

This repository contains the source code for the esphome/voice-kit XU316 microcontroller.

Building the firmware

NOTE: This guide was written using Debian Linux 12 (bookworm)

Step 1: download and install the XMOS XTC tools for your operating system.

You must create an XMOS developer account to download the tools.

On Linux:

  • Click the "Linux 64-bit" link
  • Extract the tools to a directory (example: ~/XMOS )
  • In a terminal, source the environment script (example: source ~/XMOS/XTC/15.2.1/SetEnv )
  • If done correctly, run xcc --version and it should work

Note: This process is nearly identical for macOS -- just use the macOS version of the tools and install into /Applications.

Step 2: clone the source repository.

You must have an SSH key associated with your Github account because the submodules are all referenced using SSH. If your SSH key is working, ssh -T [email protected] should succeed.

git clone --recurse-submodules [email protected]:esphome/voice-kit-xmos-firmware
cd voice-kit-xmos-firmware/
git checkout <branch>
git submodule update --init --recursive

Step 3: build the firmware.

Ensure you have at least cmake version 3.21 or higher installed.

On Linux, you also need to ensure that libtinfo5 is installed with sudo apt-get install libtinfo5

These commands must be run from the same terminal where you sourced the XTC environment script, and must be run within the voice-kit-xmos-firmware directory.

# Only run this once
cmake -B build --toolchain xmos_cmake_toolchain/xs3a.cmake
cd build/

# Run this every time (in the build directory)
make create_upgrade_img_example_ffva_int_fixed_delay

Step 4: flash the firmware.

Unplug the VoiceKit, change the switch on the VoiceKit to select the XU316 and plug it back in.

Ensure that you can see it with dfu-util (install if needed):

sudo dfu-util --list

Now flash the firmware located in the voice-kit-xmos-firmware/build directory:

sudo dfu-util -e -a 1 -D example_ffva_int_fixed_delay_upgrade.bin

Unplug the VoiceKit, switch the jumper back to the ESP32, and plug it back in.

Profit.