Jay Won Kim (Jay Kim)

Jay Won Kim (Jay Kim)

he/him

Hello! I'm an Electrical Engineering student at the University of Illinois at Urbana-Champaign with interests in Embedded Systems and Robotics. Other interests include Computer Vision as well as a hint of Psychology and History.

Pressure Piano (ECE 145)

From August 2025 to December 2025 · ECE 145 final project

A short walkthrough of the completed build (with sound).

Overview

A semester-long hardware design project: a pressure-sensitive electronic piano that generates dynamically controlled audio through fully analog and digital circuitry. Rather than simple on/off keys, it reads how hard each key is pressed and shapes the sound in real time.

The completed pressure piano: three breadboards wired together on the bench, powered from a lab supply.
The completed system — several breadboards spanning the octaves, powered from the bench supply.

How it works

  • Pressure sensing. Force-sensitive resistors (FSRs) under the keys detect finger pressure, driving real-time volume modulation and multi-key input.
  • Tone generation. 21 analog sinusoidal tone generators span three octaves (C3–B5), built from op-amp oscillators and RC networks; each was tuned and its frequency accuracy verified on an oscilloscope.
  • Control logic. Tone selection and routing run through multiplexers and decoders, with a finite state machine handling octave selection.
  • Audio output. An AD605 voltage-controlled amplifier sets output gain from the average pressure across pressed keys — harder presses play louder.
A finger pressing a key on the breadboard during testing.
Playing a key by hand during testing.
Close-up of a tone-generator board showing op-amp ICs and rows of resistors.
A tone-generator board — op-amp oscillators and the resistor networks that set each note.

Building and testing

Most of the work was hands-on mixed-signal debugging and validation with oscilloscopes and multimeters, and wrestling with noise mitigation and hardware scalability.

An earlier single-breadboard build wired on the bench.
An earlier single-board build on the bench.
A benchtop oscilloscope displaying a captured trace during testing.
Checking signals on the bench oscilloscope.

Follow-up

Following an issue noticed during the final demo, I’m currently working to address the noise that appears when the lower notes (C–E) in octaves 3 and 5 are played.