Posted: February 1, 2026

Introduction

This project is an experiment on the power output of a cheap stirling engine. I was originally wanting to do this with a steam engine, but I couldn't get the (even cheaper) steam engine kit that I got. Most likely the issue was I was trying to fire it up with a candle because lighting a container of isopropyl (rubbing alcohol) with a wick made me nervous. The same candle with the stirling engine didn't work either.

This project powers an ATtiny85 microcontroller which plays a song.

Video

The video shows first the stirling's voltage output on an oscilloscope with a capacitor to smooth it out. It then shows the output without the capacitor. Next it shows a microcontroller running on the generated electricity playing a song and finally a geared motor turning. YouTube: https://www.youtube.com/watch?v=i3b0tzk94tY

Explanation

I got this stirling engine all assembled and ready to go. It came with an electric motor which is used to generate electricity as it spins. It also had an RGB LED (probably with a small microcontroller) built into the motor and a 7 segment display that would display what appear to be random numbers. There is also a USB port on the side, which I wouldn't recommend plugging anything USB into. I removed the 7 segment display to try to route as much current as I could to the USB port so I could power the circuit.

The circuit to play the song is based on an ATtiny85. The song was programmed with Drums++ which outputs MIDI and converted from MIDI to a JSON file with parse_mid and then converted to binary data with a Python script. This process was very similar to what was done with the xylophone solenoid music project.

For sound I used two piezo elements. The idea was to keep the current draw as low as possible incase the generator couldn't put out a lot of current. It also keept the amount of circuitry down. The problem was the stirling engine is so loud it was hard to hear the music. I tried some really small speakers but the microcontroller kept shutting down, probably more due to the fact that the speakers were connected directly to the the pins of the microcontroller and starving the microcontroller of power. Since the engine can drive that electric motor, it seems it should have enough power to drive those speakers.

Source code

Copyright 1997-2026 - Michael Kohn