CoCA "Emerge + Tech" - Musical plants is an electronics, software and plants base exhibit we developed for Cinnamon Colomboscope 2016 exhibition. This installation will feature a touch sensitive sensor that will generate a soothing sound from each plant. By interacting with individual plants, the audience can create musical compositions.

Sensor unit of this project consists with 14 high sensitive, configurable level detectors and USB interface. In this sensor unit we use Samsung's KA2284 as amplifier, level detector, and PIC18F4550 8-bit MCU as data multiplexer and USB I/O controller. PC817 opto-isolators are used to interface between analog front-end and digital data processor. In this design both analog and digital circuits are physically and electronically isolated to minimize the noises and interference.

Before the KA2284 design we try several capacitive sensors with our plants installation and most of them give us a mixed results. For example MPR121 gives quiet good results with small plants, but its not much responsive with large plants like jack, costus, etc. Also we got similar results with popular TTP229 sensor IC. We also consider about Makey Makey, but the biggest problem which we got is it's limited sensitivity.

In this design we use KA2284 because it is commonly available in local market (in Sri Lanka) and costs less then 0.1 USD. To reduce the interference we use shielded cables between plants and sensor ICs. With the special wire we are using we may be able to extend the sensor connectivity up to 3m.

PC application of this project is developed by using HIDAPI and irrKlang libraries. This application is designed to work with Microsoft Windows operating systems and we tested it on Windows 7 and Windows 10.

All the audio clips used in this project are produced by using xylophone.

This project is an open source hardware project released under terms of MIT and CC BY-SA licenses. All the content of this project are available to download at https://github.com/dilshan/musical-plants. Conceptual details of this project is available at CoCA homepage.