COMP3601 - Design Project A
22T3
FPGA-driven Microphone System

In this project, you will be designing a custom microphone system using a multiprocessor
system on a chip (an MPSoC) system that has an FPGA and ARM microprocessor (assume
you are a start-up company who wants to develop custom audio hardware and
applications). Your product involves capturing user audio (as a personal recording device)
via a microphone and then processing the audio using the MPSoC.

There are two parts/tasks to the project with the first task being to develop a basic
microphone system and the second task extending the basic design based on each team’s
choice of extension. The basic microphone system will include an I2S MEMS microphone
breakout board interfaced with the FPGA. The FPGA module of the system will drive the
microphone, collect the audio samples from the microphone, and send the audio samples to
the software running on the ARM processor. As an example, the software for the basic
system can be a simple audio recorder that records and save the collected audio samples
into a wav file that can be played back using a media player or into a csv file that can
opened with an application such as MATLAB to display the waveform or playback.

The second task will be a more open-ended system extension task where teams can decide
what features to add to their microphone system. See the deliverables and talk to your tutors
for what options can you extend on.














Xilinx Kria KV260 Board
I2S MEMS Mic.
Audio
Samples

Terminal Wave file
Development Board –
The development board (MPSoC Board) chosen for this course is the Xilinx Kria KV260 Vision
AI Starter Kit which is a development kit that includes a Zynq UltraScale+ MPSoC. The MPSoC
IC combines a quad-core ARM Cortex A53 hard processor (usually referred as processing
system or PS) and an FPGA (referred as programmable logic or PL). PS will run a custom Linux
image which referred to as PetaLinux. On the top of PetaLinux, drivers and recorder
application (or any other application) can be executed to save audio samples from the
microphone (or perform any software operation that you intend).

I2S MEMS Microphone
I2S MEMS Microphone is a small microphone that converts sound to voltage and gives out
the sampled audio as a purely digital signal. This microphone is capable of capturing sound
waves with frequencies ranging from 50Hz to 15KHz, which is good for all general audio
recording/processing applications. The microphone supports I2S (Inter-IC Sound) serial bus
interface protocol, which is used for connecting digital audio devices together, and this
protocol is being used to transfer the sampled audio data (digital) to the FPGA-based audio
processing pipeline in the project.


Source Code Structure
Components Provided
• 1x Kria KV260 Vision AI Starter Kit
• 1x Power Supply
• 2x I2S MEMS Microphone
• Wires

Deliverables
Task1:
Design and develop a basic I2S microphone system using the provided I2S MEMS microphones
and the Kria development board. The system should be able to record an audio clip from the
microphone that are wired to the PMOD pins connected to the FPGA of the MPSoC.
Task2:
Extend on your basic I2S microphone system. This could be:
• Electronics
o Supporting microphones of other protocols
o Adding physical control methods to the audio
o Others
• FPGA
o DSP (Volume/Gain control, FIR filter, etc.)
o Audio monitoring
o Audio format (sample rate, etc.)
o Others
• Software application
o Audio messaging
o Audio monitoring
o Audio networking
o Others
Be curious and use your creativity for task 2.

References
• https://www.xilinx.com/products/som/kria/kv260-vision-starter-kit.html
• https://www.xilinx.com/products/silicon-devices/soc/zynq-ultrascale-mpsoc.html
• https://www.adafruit.com/product/3421
• https://digilent.com/reference/pmod/pmodbb/reference-manual?redirect=1
• https://en.wikipedia.org/wiki/I%C2%B2S