A shared instrument with wooden keys
for resource-limited classrooms.
Independent BA graduation project · Feb–May 2025
Material experiments, product design & a working touch-and-sound prototype.
Named on a related patent application accepted by CNIPA.
2026 extension: product model & classroom concept.
The instrument starts with touch.
The project starts with access.
Wood, colour and sound in one shared instrument.
How can children in rural China take part in music with fewer specialist teachers and classroom resources?
In 2019, 30.0% of rural primary-school heads said their music-teacher numbers met teaching needs, up from 14.5% in 2016. In city schools it was 70.1%.
So the instrument is shared. One instrument for a group, five notes on one controller and one sound output.
Figure 14 · primary-school heads
In 2019, 41.7% of Grade 4 music teachers taught music full-time. Nearly a third, 31.7%, taught it part-time without related training.
So it has to be easy to start. Colour and number instead of notation, and a one-page activity for the teacher.
Table 4 · Grade 4 music teachers, national
Student instruments had the widest gap in Grade 4: a provision rate of 78.2%, while only 40.6% of music teachers in equipped schools used them often.
So it has to stay in use. Quick to set out, sturdy to share, and repairable: a wooden face lifts out, the parts are reached through the base.
Table 2 · usage = arts teachers in equipped schools who use it often
| Place | 2016 | 2019 |
|---|---|---|
| City | 54.9 | 70.1 |
| County town | 38 | 42.8 |
| Township | 26.1 | 38.7 |
| Rural | 14.5 | 30 |
| Grade | Full-time | Part-time, trained | Part-time, untrained |
|---|---|---|---|
| Grade 4 | 41.7 | 26.6 | 31.7 |
| Grade 8 | 78 | 12.6 | 9.3 |
| Resource | Grade 4 provided | Grade 4 used | Grade 8 provided | Grade 8 used |
|---|---|---|---|---|
| Music classroom | 82.4 | 64.1 | 90.5 | 71.1 |
| Teaching aids | 90.9 | 59.2 | 95.9 | 69.7 |
| Student instruments | 78.2 | 40.6 | 78.3 | 36.5 |
| Sound system | 89.5 | 73 | 91.7 | 82.4 |
The 2025 personas named three needs each. Choose one to see where the instrument answers it.
Each need maps to one part of the instrument.
All three figures come from the 2019 National Compulsory Education Quality Monitoring, arts results, published by the Ministry of Education in 2020: Figure 14, Table 4 and Table 2 ↗. Figure 14 reports what primary-school heads said about music-teacher numbers. Table 4 covers Grade 4 and Grade 8 music teachers nationally. The usage rate is the share of arts teachers in schools with the resource who use it often. The report's rural arts-teacher statistics exclude schools with fewer than 100 pupils.
The needs come from the two personas in the 2025 graduation project. They are design inputs, not test results.
A wooden touch surface, grown out of a material experiment.
A natural face to touch.
A dark, processed face underneath.
A natural face to touch. A treated and laser-processed surface underneath.
Processed face
Laser alone only browns wood. Coated with an iron–tannin ink first, the surface turns into porous, graphite-like carbon that conducts.
up to 2,500 S/mone laser passRaman D ≈ 1350 · G ≈ 1575 cm⁻¹
Dreimol et al., Nat. Commun. 2022, fig. 3f, g · CC BY 4.0, cropped. Reference study, not these samples. Raman spectroscopy ↗
Natural wood face
The side a child touches stays natural wood. In the reference study only the coated surface converts; the wood beneath keeps its structure.
veneers ~0.4–1.5 mmsurface layer only
Dreimol et al., Nat. Commun. 2022, fig. 3a · CC BY 4.0, cropped.
Local contact
A small contact links the processed face to its lead. The reference study wired each engraved area to an Arduino with an MPR121 board, as Resonance does: touching the natural face on the other side changes the capacitance.
self-capacitanceArduino + MPR121
Dreimol et al., Nat. Commun. 2022, fig. 5e · CC BY 4.0, cropped. Capacitive sensing ↗
Recipe · 2025 report, fig. 3-17
50 ml water + 50 ml persimmon lacquer
150 ml water + 1 g green vitriol
Tannins bind iron into a dark iron–tannin complex, the chemistry of iron-gall ink. The reference study used tannic acid, iron citrate, gum arabic and glycerol.
First test · not conductive
Treated wood, no laser yet
Wood does not conduct on its own, and the dark colouring is not yet carbon. In the reference study, ink without laser showed no graphite-like signal.
Laser settings
2025 trial: 250 mm/s
Reference: ~13 W at 200 mm/s, or ~12 W at 270 mm/s
The reference study used a 10.6 µm CO₂ laser (60 W maximum) in one pass, in ordinary air.
What changed
Thicker boards
Adjusted ratio
Thin boards and bamboo strips gave unstable results, so the ratio was adjusted and thicker boards were tried. Durability is still to be tested.
The reference study used iron–tannic-acid ink to promote laser-induced graphitisation. My trials explored persimmon lacquer and green vitriol. Dreimol et al., 2022 ↗
Material experiments were supported by Wei Sentao. Original photographs and trial sequence: 2025 graduation report, figures 3-17 to 3-24.
The 2025 prototype linked the processed wood to an Arduino UNO, an MPR121 touch sensor and a VS1053 audio board. The 2026 model keeps that chain inside one shell. Move the lens to look inside, then touch a key.
key 3 · Miif (touched & (1 << 2))noteValues[] = {60, 62, 64, 67, 69}audio.midiNoteOn(0, 64, 127);
// 516.ino · MPR121 five-point test (I²C address 0x5A)
uint16_t touched = cap.touched();
for (uint8_t i = 0; i < 5; i++) {
if (touched & (1 << i)) { /* input i touched */ }
}
// touch-11.ino · VS1053 in MIDI mode
int noteValues[] = {60, 62, 64, 67, 69}; // C D E G A
audio.useMIDI();
audio.midiNoteOn(0, midiNote, 127);Excerpts from the two test sketches in the 2025 project files. The final prototype combined MPR121 sensing with VS1053 sound on an Arduino UNO R3.
From a pentagon of five modules in 2025 to one shared instrument in 2026. Scroll, or drag the lens along the track.
2025 · Form
Form studies settled on a pentagon. Each side would carry one sound channel, around a shared centre.
2025 · Drawings
The Rhino model and its drawings fixed the module sizes. Renders checked there was room for the Arduino and the touch sensor.
2025 · Render
Colour joined the wood: each module took the colour of its note, with a wooden top to touch.
2025 · Prototype
Shells were SLA-printed in white and translucent resin. The UNO and its shield sat in the centre, with leads out to each module.
2026 · Instrument
The five notes now sit in a single row on one shell, easier to share across a table. The electronics have one home, the wooden faces lift out, and the parts are reached through the base.
The graduation project explored separate modules around a centre plate. These original boards show the starting point for the later shared enclosure.
Opens the full-resolution board in a new tab 
Opens the full-resolution board in a new tab 
One instrument on each group table, an activity sheet and three role cards: a short guided activity the teacher can begin with.
Three children share one instrument and one activity sheet. After each short phrase, the roles move one seat round the table.
PlayComposeListen Concept visualisation · one group table, seen from above Play a short phrase. The next person repeats it.
Phrase3 3 5 6 → 3 3 5 6
Choose a different note or rhythm. Hear the difference.
Phrase3 3 5 6 → 3 3 5 1
Agree a phrase, play it together, then swap roles.
Phrase3 3 5 6 i i 6 5
Place one instrument and the activity sheet on each group table. Demonstrate a short phrase, then let the children take turns.
Download the activity sheet Arts curriculum, 2022, pp. 15–16 ↗
A shared instrument on each table, teacher guidance across the room, and space for children to listen and create.
The instrument needs to be easy to begin with, easy to share and practical to keep using. A material experiment became a wooden touch surface, then a shared instrument and a short classroom activity.
