How handpans are made

MAK-01How is a handpan made, from flat sheet to finished instrument?#

beginner

Short answer. A flat steel disc is formed into a shallow bowl, tone fields are marked and hammered into it, the shell is heat treated, then tuned by hammer over many sessions with rest periods between them, and finally glued to a second bowl that carries the sound hole.

The sequence almost every maker follows, whatever their individual methods:

1. Blank. A circular disc of sheet steel, typically somewhere around 0.8 mm to 1.2 mm thick and 45 cm to 60 cm across, is cut or bought ready cut.
2. Forming. The disc is turned into a shallow dished bowl, either by hammering it down by hand, by pressing it, by spinning it on a lathe, or by hydroforming it with water pressure. See MAK-02.
3. Marking out. The tone field positions are laid out on the top shell, usually from a template the maker has developed over years.
4. Rough shaping. Each note gets its dimple and its surrounding border hammered in. This is where the note is born, and it is mostly rough work.
5. Heat treatment. The shell is heated and cooled under control. For carbon steel this often means nitriding, a long furnace cycle described in MAK-03. Others anneal, temper or normalise instead.
6. Tuning. The maker works each note towards its target frequencies with a hammer, checking constantly against software or a strobe. This takes many passes.
7. Resting and retuning. The instrument is put aside for days or weeks, then brought back and corrected. Repeat until it stops moving. See MAK-06.
8. Assembly. Top and bottom shells are glued together, the rim is trimmed and finished, and the surface is cleaned and protected. See MAK-07.
9. Final check. A last tuning pass after assembly, because joining the shells changes the sound, then playing in, photographing and packing.

Different makers reorder steps 4 to 7 substantially, and some run heat treatment more than once. What almost nobody does is a single linear pass. The instrument goes forwards and backwards through shaping, heating and tuning until it settles.

See also: MAK-09, Anatomy and construction

MAK-02How are handpan shells formed, and does the method change the result?#

intermediate

Short answer. Four methods dominate: hand sinking, deep drawing, spinning and hydroforming. They differ mainly in how evenly the steel thins and how much internal stress they leave, and makers argue about whether the audible difference is large or mostly marketing.

Hand sinking is the traditional steelpan approach. The maker hammers the disc down into a dish over thousands of blows, working from the outside in. It thins the metal unevenly, work hardens it, and leaves a distinctive hammered texture. Advocates say the resulting stress pattern gives a warmer, more individual voice. Critics say it makes consistency between instruments almost impossible.

Deep drawing presses the disc into a die in seconds using an industrial press and a blank holder ring. It is by far the most common source of shells today, because it gives very even wall thickness and repeatable geometry. Several specialist suppliers sell deep drawn shells to makers, which is a large part of why the number of handpan makers grew so quickly after about 2010. A maker buying shells is not necessarily cutting corners: the shaping and tuning that follow are still the hard part.

Spinning turns the disc on a lathe while a roller or tool progressively forms it into a bowl. It produces a smooth, even shell and was used by some early makers as an alternative to hand sinking.

Hydroforming uses water or hydraulic pressure to push the sheet against a die or simply into free form. It is gentle and even, and it can be done at surprisingly modest cost. Colin Foulke publicly demonstrated a version using a pressure washer, frames and bolts, which was influential in showing hobbyists that shell forming was not out of reach.

Does it change the sound? Honestly, the forming method matters less than what happens afterwards. Wall thickness distribution and residual stress do affect how a note behaves under the hammer, and a hand sunk shell behaves differently from a drawn one in the tuner’s hands. But two instruments from the same maker on different shell types usually sound more alike than two instruments from different makers on identical shells. Be sceptical of any claim that one forming method is inherently superior.

See also: MAK-01, MAK-03

MAK-03What is nitriding, and why do makers do it?#

intermediate

Short answer. Nitriding is a long, hot furnace treatment that drives nitrogen into the steel, forming hard iron nitrides. It makes the steel stiffer, more corrosion resistant and much better at holding a tuning, at the cost of some brightness and sustain if it is overdone.

The commonest version in handpan making is gas nitriding: the shell sits in a furnace in an ammonia atmosphere, typically somewhere in the region of 490 to 580 degrees Celsius, for anything from a few hours to well over a hundred. The ammonia dissociates and nitrogen diffuses into the surface, forming a hard compound layer and a deeper diffusion zone of needle shaped iron nitrides.

Plasma or ion nitriding does the same job using an electrically excited gas at lower pressure. Makers who have compared the two often report that plasma nitrided instruments come out brighter, with less damping of the high partials, while gas nitriding tends to stiffen and slightly mute the sound.

Three reasons makers bother:

• Rust. Handpans live under salty, acidic human hands. Untreated carbon steel corrodes fast. Nitriding is not stainless, and a nitrided pan will still rust if neglected, but it buys a lot of margin.
• Tuning stability. Steel has a memory. Hammered tone fields want to creep back towards their previous shape. Hardening the metal resists that creep, which is one of the main reasons nitrided instruments hold tune better than untreated ones.
• Tone. The stiffening changes the timbre. Some of the character people associate with a good handpan comes from this treatment rather than from the tuning alone.

The trade off is real. Push the treatment too deep and you get an instrument that is very stable and very dull, with short sustain. Keep it light and you keep brightness but gain little rust protection. Every maker who uses the process describes their furnace recipe (temperature, hold time, gas mix, pressure, single stage or two stage) as something they refined over years, and most will not share it.

PANArt held a patent on an “exhaustive” nitriding method that nitrided through the full sheet thickness rather than just the surface, filed in June 2009 and granted at the EPO in 2014. The European patent was revoked in 2023, and the family lapsed in several jurisdictions from 2020 for non payment of renewal fees.

Going deeper. Nitriding is done after shaping but usually before final tuning, because the hardened steel behaves differently under the hammer and a maker needs to tune the material as it will actually be. Some makers run more than one furnace cycle across a build.

See also: MAK-14, Rust and corrosion

MAK-04How are the tone fields laid out on the shell?#

intermediate

Short answer. The maker marks note positions from a template developed over years, sizing each field to its pitch (bigger and deeper for lower notes) and arranging them so that neighbouring notes are not intervals that will make each other buzz.

Layout is the step that looks trivial and is not. It takes minutes of physical work and years of knowledge, because once the fields are hammered in, the geometry is fixed.

Several things have to be reconciled at once:

Size against pitch. A tone field’s fundamental depends chiefly on its area, the curvature of its dimple, and the stiffness of the border around it. Low notes need large, deep fields; high notes are small and shallow. The lowest note of the scale, the ding, sits in the centre as a dome and behaves differently from the rest.

Fitting the scale on the shell. A nine note instrument on a 53 cm shell has plenty of room. A fourteen note instrument on the same shell does not. Fields have to shrink, borders get thinner, and the maker starts trading sustain and volume for note count. This is a real reason why very high note count instruments are harder to build well and often sound tighter.

Crosstalk. Adjacent fields share the same sheet of steel and couple through it. Put two notes an octave apart side by side and striking one will visibly excite the other. Some coupling is desirable and part of the instrument’s bloom. Too much and you get sympathetic ringing that muddies everything. Makers manage this by choosing where in the circle each pitch sits, by spacing, and by how they shape the borders. Standard handpan layouts, with the scale alternating left and right up the circle, evolved partly for playability and partly because it separates the intervals that fight.

The bottom shell. The gu, the tuned port in the lower shell, is a Helmholtz resonator. Its size and the shell volume set its pitch, and it interacts with the lowest notes.

Makers guard their templates. A template encodes the answers to hundreds of small questions about a particular shell size, steel thickness and scale, and copying one is copying much of the maker’s accumulated work.

See also: MAK-05, Scales and music theory

MAK-05How is an individual note shaped?#

intermediate

Short answer. The maker hammers a shallow concave dimple into the marked area and works a raised border around it, which sets the field’s boundary condition. The dimple and the border together determine the note’s pitch, its overtones and how cleanly it speaks.

Two features do the work.

The dimple is the shallow depression in the centre of the field. Hammering it in stretches and thins that patch of steel, and the resulting curvature and tension set the fundamental. Deeper and wider generally means lower.

The border is the ridge or transition worked around the field’s edge. It is the boundary condition for the vibrating area, and it matters enormously. A well defined border isolates the field, gives a clear pitch and reduces bleed into neighbours. A vague border gives a woolly note that leaks energy sideways. Makers describe border work as having as much effect on the note’s character as the dimple itself.

Rough shaping is done with heavier tools: air hammers, pneumatic hammers, or hand hammers on a form. Some makers use presses or dies for consistency, others insist on hand work because the slight variation from blow to blow gives each note an individual character. Both positions are defensible and both produce good instruments.

At this stage the note is nowhere near in tune. It is roughly the right size, roughly the right shape, and roughly the right pitch. The point is to establish geometry and to get the steel work hardened and stretched into a stable configuration. Everything after this is refinement measured in single cents.

Going deeper. The three main partials respond differently to where the hammer lands. Broadly, work in the centre of the dimple moves the fundamental most, work around the border and the outer parts of the field affects the higher partials more, and work outside the field affects the tension holding everything. This is the lever a tuner uses to move partials independently, and it is covered in MAK-11.

See also: MAK-11, Acoustics and physics

MAK-06Why is a handpan tuned, rested and retuned over and over?#

intermediate

Short answer. Hammering leaves the steel under internal stress, and that stress relaxes over hours and days, pulling the notes off pitch. Tuning, resting so the steel can move, then correcting is the only reliable way to reach a tuning that stays put.

Think of a tuned tone field as a balance of tensions in a thin sheet. Every hammer blow adds local stress. Immediately after a session the instrument can measure perfectly in tune and be lying to you: over the next days the stresses redistribute, the field creeps, and the pitch drifts. Play it hard and you accelerate the same process.

So the workflow is cyclical. A maker will typically:

• tune a pass, getting everything close
• put the instrument aside, from a couple of days to a couple of weeks
• come back and measure what has moved
• correct, often with much lighter hammer work
• rest again
• repeat until a rest period produces almost no movement

Some makers add heat between passes: a controlled bake or anneal that relieves stress deliberately rather than waiting for it to relieve itself. Others tap or play the instrument hard on purpose to provoke movement before the customer can. Either way, the goal is the same, which is to get the steel to do its settling in the workshop instead of in the buyer’s living room.

Amateurs almost always underestimate this. A common piece of advice from experienced builders to beginners is to leave three or four days between heavy tuning sessions and to expect fine tuning done immediately after rough work to be wasted effort.

How many passes? It varies enormously. Some makers report three or four cycles; others take an instrument through many more over weeks or months and treat the wait time as the single biggest difference between a stable instrument and one that drifts within a year. Makers who rush this stage produce pans that measure well on the day of sale and need retuning surprisingly soon.

See also: MAK-14, Tuning stability

MAK-07How are the two shells joined?#

intermediate

Short answer. Almost all modern handpans are glued, not welded, using a flexible industrial adhesive along the rim. Welding is avoided because the heat would ruin the tuning and the rigid joint would damp the instrument.

The two formed shells meet at the rim, and how that joint is made has real acoustic consequences. A rigid joint couples the shells strongly and tends to kill sustain. A joint with some compliance lets the top shell vibrate more freely while still forming the sealed cavity the instrument needs.

The standard solution is a flexible polymer adhesive. One product, Merbenit HS60, has been used widely in the trade for well over a decade and is sold by handpan supply shops alongside shells and tools, which tells you how settled this part of the craft is. The adhesive is run around the rim, the shells are clamped or strapped, and the assembly is left to cure.

Variations you will encounter:

• Rim treatment. After curing, the excess is trimmed and the rim is sanded, filed or polished. Some makers leave a visible glue line; others finish it almost invisibly. This is cosmetic, not acoustic.
• Riveting or bolting. A few builders add mechanical fasteners, usually on larger or unconventional instruments.
• Welded or brazed rims. Rare in handpans, common in some steelpan traditions. It makes later repair and retuning far harder.
• Bolted or removable joints. Occasionally used experimentally so the interior can be accessed, but they tend to buzz.

Joining changes the sound. Sealing the cavity brings the gu resonance into play and slightly shifts the pitches of the notes, particularly the low ones. That is why most makers do a final tuning pass after assembly rather than treating gluing as the last step.

The joint is also a maintenance point. A knock on the rim can crack a glue line, and instruments do occasionally come apart at the seam, usually after impact or extreme heat. It is repairable by a maker.

See also: MAK-01, Parts of a handpan

MAK-08How many instruments do makers scrap, and why?#

advanced

Short answer. Nobody publishes reliable figures, but a loss rate somewhere in the range of roughly one in ten to one in five is often quoted for established makers, and far higher for beginners. Instruments are scrapped for cracks, for notes that will not stabilise, and for tone the maker will not put their name on.

Treat any precise number with suspicion. Makers have no incentive to publish scrap rates, and the figures that circulate come from interviews and blog posts rather than audited data. What is clear is that the rate is high enough to be a major cost driver, and that it falls steeply with experience.

The common causes:

Cracking. Thin steel that has been repeatedly stretched can tear, usually at a border or at the edge of a dimple. Once a tone field cracks it is effectively finished, since the crack changes the boundary condition and will propagate. Cracks are more likely on aggressive shaping, on thin steel, on high note count layouts, and on notes pushed far from the size their pitch wants.

Notes that will not settle. Sometimes a field simply refuses to hold. It goes in tune, rests, drifts, gets corrected, drifts again. At some point the maker decides the steel in that spot is too worked to hold anything and the instrument comes off the bench.

Partials that will not line up. Getting the fundamental right is achievable. Getting the fundamental, octave and compound fifth all right at once, on every note, without wrecking a neighbour, is where builds die. A note where the octave is stubbornly flat while the fundamental is correct may not be fixable.

Aesthetic and tonal rejection. Plenty of scrapped instruments would play. They are scrapped because the maker judges the sustain too short, the timbre uneven across the scale, the crosstalk excessive, or the finish poor. This is a judgement call, and where a maker draws that line is a large part of why prices differ between workshops.

Some makers sell rejected builds as seconds, clearly labelled and heavily discounted. Others destroy them rather than let anything below standard carry their name. Neither approach is wrong, but as a buyer it is worth knowing which one you are dealing with.

See also: MAK-09, Prices and value

MAK-09How long does one handpan take to make?#

beginner

Short answer. Perhaps forty to eighty hours of actual work per instrument for an experienced maker, but spread over weeks or months of calendar time because of the resting periods between tuning passes.

There are two clocks running and people confuse them constantly.

Hands on hours. Forming, shaping, tuning passes, assembly and finishing. Estimates from makers commonly land in the range of forty to eighty hours for a standard instrument, with tuning taking the majority of it. The number climbs sharply for high note count instruments, unusual scales, or anything the maker has not built before.

Calendar time. Because the steel has to rest between tuning sessions, and because heat treatment is often outsourced to a furnace with its own schedule, an instrument that represents fifty working hours may take six weeks to three months to pass through the shop. Makers with long waiting lists are usually not slow at the bench; they are batching instruments through a process with unavoidable waiting built into it.

This explains two things buyers find frustrating. First, waiting lists of six months to two years at well regarded workshops are normal rather than a sign of disorganisation. Second, prices do not fall much with volume, because there is no meaningful economy of scale. Every instrument is shaped and tuned individually by a skilled person, there are no interchangeable parts, and the hardest step cannot be automated. Most respected small workshops turn out something in the low hundreds of instruments a year at most, and many make far fewer.

A useful sanity check when you see a very cheap instrument advertised for immediate dispatch: ask yourself how many hours of skilled tuning the price can possibly contain. Sometimes the answer is that the shell was bought in and the tuning was quick, which can still produce a serviceable instrument. Sometimes the answer is that nobody really tuned it.

See also: MAK-08, MAK-19, Makers and ordering