Hulam for Celesta

The second very important part of the instrument is the sound source. In our case, these are the Hulam pipes. I don’t write the articles in the order the individual parts were created – in fact, Hulam came first. I needed to check whether the individual tones would even be audible and, at the same time, determine a usable range. Originally, I experimented with aluminum pipes, but they didn’t have a long enough sustain, so brass came into play. That was a whole different story – I’m still fascinated by the endless resonance of brass tones.

Another thing I had to change compared to the Hulam, which I usually make, was the arrangement of the tones and resonators. The resonators have a larger diameter than the pipes, so it was necessary to choose dimensions that would allow the individual tones to fit next to each other. I tested these proportions on the piano we have at home. All of this happened before I found a donor, so I first determined the maximum possible diameter of the brass tone pipe, compared it with available sizes, and chose the right solution accordingly.

Experimenting on our piano
I also tried an aluminum version.

Since the resonators have a larger diameter and couldn’t fit side by side, I had to position them perpendicular to the tones and alternately.

One of the boards holding the tones and resonators. Each hole in the board corresponds to a single resonator, and therefore to a single tone.

A resonator is a tube tuned to exactly the same frequency as the brass tone. This allows it to amplify the tone through a phenomenon called resonance, which is where its name comes from – resonator. The deeper the tone, the longer the resonator.

What sets the Hulam celesta apart from other celestas is its unique use of open resonators and tubular tones without any additional modifications. Standard celestas typically use metal plates for the tones and closed resonators. Moreover, the plates are often fitted with weights to artificially lower the pitch while keeping the original size.

I avoided this, which allows me to achieve a very long sustain.

Each tone needs to be drilled in two places. The placement of the holes must be extremely precise (to a tenth of a millimeter), otherwise the sustain is significantly shortened.
Each hole needs to be thoroughly cleaned. Burrs and drilling residue can easily wear through the wire, causing the entire assembly to fall apart.
Final cleaning of the holes.
Each tone needs to be tested several times.
After tuning each tone and preparing the board, assembly could begin. I threaded each tone onto a string, then tightened and tied the string at the end.
The top four octaves of the instrument, temporarily assembled. The lower tones required bends so the pipes wouldn’t protrude too far from the instrument (the limit was 80 cm depth to fit through doors). Three tone boards are mounted on the instrument (based on the division of hammers in the mechanism).
Prototype boards—the largest one was later modified so the tones would stay in place and sound as intended. The second lowest octave is missing in the picture.
The second lowest octave added. With this range (four octaves), the instrument was already usable for a wide variety of pieces.
Of course, I was impatient, and with every step forward I needed to play and test the instrument.
One of the tests. Everything was temporarily assembled, including the pedal mechanism…

The bass tones—the lowest, fifth octave—were a special category. I deliberately left them for last, because until the very end I didn’t know if or how I would assemble them. In any case, I knew I wanted them in the instrument.

I started by testing the tones. I used aluminum pipes, the same ones I usually use for Hulam tones. Brass tones would have been three times heavier, and at this size, even aluminum had a good sustain.

Testing the instrument’s lowest tone (C3).

These tones, of course, could no longer fit side by side above the keys like the other octaves, so I had to come up with a more original solution.

I arranged the tuned tones as close together as possible.
At the same time, I tested how the tones would fit into the prepared space in the instrument. The resonators for the bass and other tones are not centered but positioned at the edges. They couldn’t be centered because of the dimensions. The sound, of course, was not affected—the pipe vibrates just as well at the edges as in the center.
The bass assembly placed in the instrument. Next to it, the large hammer for sounding them is already prepared. The mechanics of the bass section will be covered in a separate article.
Part of the resonators is placed beneath the tones. Since they are larger than the tones themselves, a resonator fits under every other tone. The remaining ones had to be routed above the mechanism.
The completed bass assembly with the lower resonators.