Building a Trumpet

Finished brass horn

In the spring of 2024 I took ME45: Musical Instrument Design at Tufts, where I set out to build a trumpet from scratch.

Building It:

I started with the mouthpiece. I'd planned to speed things up by turning it on the CNC lathe, but ended up hand-turning it on a regular lathe instead, cutting it down from a stainless steel cylinder — which turned out to be way more fun than the CNC version would have been. It also took a lot longer than I expected: over a month just for the mouthpiece.

Hand-turning the mouthpiece on the lathe My mouthpiece next to a real one

The mouthpiece I made on the lathe is on the right.

The bell gave me the most trouble. My first plan was to CNC-cut a wooden mandrel out of plywood on the ShopBot, glue it together, and hammer brass around it to form the shape — the mandrel itself took an afternoon to build and came out looking pretty good, but it broke the moment I started hammering brass around it. Free-handing the bell without a mandrel didn't go much better; my first attempt looked rough enough that I was fairly discouraged.

Wooden bell mandrel First failed bell attempt

The breakthrough came over a weekend at home. On advice from a few classmates, I found that brass shapes just fine if you hammer it hard enough — no heat needed. I'd been trying to make a piece to extend the bell's throat, and realized that piece was actually already the right size to be the bell itself. Back at school, I cut a donut-shaped ring of brass sheet and shaped it around a leftover mandrel piece to make the flare, then learned to solder brass (flux and a torch) to join the flare to the rest of the bell.

Finished bell

At that point I decided to build a horn rather than a full trumpet — building valves felt like an entire second project on its own. I'd originally wanted to add a tuning slide so I could play bugle calls on it, but bending brass sheet into curved tubing turned out to be hard enough that I just soldered the whole thing together instead (the learning here was that I should have ordered brass tubing instead of trying to create my own tubing from brass sheet metal). I made the straight tubing by wrapping brass sheet around a plastic rod sized to the inner diameter I wanted, then tried bending it on a pipe bender — which promptly destroyed a couple of tubes, since my tubing's outer diameter was too large for it. I ended up making a second, smaller batch of tubing just for the bent sections. Soldering it all together went fairly smoothly since I'd already gotten the hang of it on the bell.

Tubing laid out before soldering Finished horn

Testing It:

My hand-turned mouthpiece sounded close to a real one on its own, just noticeably harder to hit higher notes on. The bell by itself (no mouthpiece) played a clean, clear note — about an Ab4 — with obvious harmonic overtones showing up in both an STFT and a power spectrum, and it genuinely sounded like a horn.

Bell STFT and power spectrum

Once I put the mouthpiece and the rest of the tubing together, though, it didn't sound like a horn at all — more like a buzzy, slightly amplified mouthpiece, and I could really only play notes in the first octave. My best guesses for why: the mouthpiece probably wasn't seating correctly in its receiver, the tubing was narrow with too many size transitions and kinks to resonate cleanly, there was a fair amount of stray solder inside the tubing from joining everything together, and there were likely still a few unpatched holes. Realistically, it was probably some combination of all of that.

I'd also planned to add a rim to the bell, but after running an experiment with a classmate, Rose Kitz, to see whether the added mass changed the horn's harmonics, we found no noticeable difference — so I left the rim off.

← Back to projects