Blog - Sketching with Hardware

Team 3: MIDI Panflute

Published on: | Author: Fabian Nußberger | Categories: 2015a, Best Projects, Projects

Team 3 consisted of Fabian and Alex, two “Medieninformatik” Master students. Since we have worked together on many projects before, we decided to team up with other people this time. But fate had other plans and be both drew the number 3 when teams were assigned. And so the dream team was at it once again!

Keyboard hacking:
On the first day we took our first steps in the world of electronics. After a crash course we made our first keyboard hack in the form of a game controller for the classic SNES-Game Bust a Move. The controller matched the targeting system in the game and allowed for unprecedented levels of immersion.

The Concept:
In a world, where almost every classic instrument from the guitar to the piano, has been digitized we found one instrument that has not yet made the transition to bleeps and bloops: The panflute. Coming from South America and spreading all over the worlds pedestrian areas, this classic wonderful instrument is sorely missing from the arsenal of electronic musicians.
We saw a world without a MIDI-Panflute and said : This will not do!

A midi panflute

Behold the MIDI panflute.

The panflutes blueprint.

The panflutes blueprint.

Our main goal for the flute was, that each pipe would produce a different tone when blown.
To achieve a full range of 8 tones we placed a light sensor on top of each pipe that would modulate the tones when obstructed by a finger.

The produced tones can be “piped” via Midi out to other instruments.

The Silence of the Mice/Harvesting of Mice
We pondered how we could detect when, and maybe how hard, someone was blowing into one of the pipes. Our idea was a LED coupled with a small light sensor and a small propeller in each pipe. When air is moving the propeller it interjects the LEDS lights path towards a photo resistor , which then sends the necessary data to the arduino microcontroller.

Get Pipe!

Get Pipe!

After a tip by our tutor Bernhard we went looking for old ball-mice, which house exactly what we needed: infrared based rotary decoders. We found two of them hiding in Fabian’s basement, so we decided to build four pipes, as each mouse has two sensors: one for the x-axis and one for the y-axis. We headed to the next hardware store and bought 3 meters of pvc pipes for our flute.

Harvesting a mouse for vital sensors.

Harvesting a mouse for vital sensors.

After a lot of experimental soldering and coding with no available documentation, we got the sensors, which we harvested from the first mouse, to detect rotation.

The first handmade "Rotorschlitten" prototype, still using mousesensors

The first handmade “Rotorschlitten” prototype, still using mousesensors

And so it begins anew
Unfortunately, when we opened the second mouse, we found a completely different set of electronics. Since getting the first mouse to work took so long, we decided to scrap the idea of reusing old hardware and to build the sensors ourselves. So over the weekend, we bought new photo resistors and LEDs from the electronics store. We tested them on a breadboard and prepared the code logic as far as possible, so that we would be able to finish in the last two days.

Constructing the Flute
To avoid cutting the pipes in two or complicated construction inside the pipes a “sled” was designed that would hold all necessary sensors and propellers, which could then be completely assembled and slid into the pipe. This design was lasercut, assembled and soldered.

Testing the "Rotorschlitten" sensors.

Testing the “Rotorschlitten” sensors.

With the help of a wide array of tools, we connected the pipes and cut of the ends to give our instrument the typical panflute-like shape with the sloped endings. Now it was time for testing. In this place we need to thank the “Magic Mashroom” team which provided the first device, which could be hooked up to our flute and produce some sweet sweet music. Surprisingly our flute worked almost perfectly right out of the gate.

Finishing touches and the Orange LED Incident
Meanwhile we still had problems with the light sensors. The sensor would always fire when the user would move the instrument and the light condition changed. Our first solution was a button which would manually reset the thresholds that would have to be passed for a sensor to fire.
Later we hacked together a code based solution, that calculates the average threshold and resets them accordingly.

Things are starting to take shape.

Things are starting to take shape.

After finishing up trimming and resoldering the wires to make the device actually mobile and handheld around nine in the evening, we stumbled upon a new feature. One of our green LEDs was suddenly glowing orange.

One LED is out of order and will soon die..

One LED is out of order and will soon die..

After finding a youtube video with the title “Green LED shines orange and soon dies”, panic mode was initiated and we learned, that if you decide to mark your wires you should do it correctly. What followed was the resoldering of the world’s shortest wire.

With the Arduino strapped to its back, the pipe was ready to rock:

Final Words

Sketching with Hardware was a great experience. It took our fear of working with electronics and showed us how simple it can be. Making real things move, not software, really is fun.

Taking apart old hardware and using it was an interesting aspect of the course. For a project on a tight schedule like this one I would suggest making your own designs, because a hardware hack like our “stolen” mouse sensors just might not work the way you want it to.

linked categories 2015a, Best Projects, Projects


Write a comment