Every headset owner knows the pattern: the drivers still sound fine, the microphone still works, the cable is intact, and yet the whole thing is headed for the bin because one small plastic joint between the headband and an earpiece has snapped. Hackaday writer Jenny List hit that wall again with her latest headset and, rather than simply accept it, designed a replacement band in OpenSCAD. She describes the project in A Headset Fit For A Hackaday Writer, published October 6, and has posted the source as jennys-headset on GitHub under a CC-BY-SA-4.0 license. Its central idea, replacing a mechanical joint with a printed zig-zag spring, is useful well beyond headsets.
Three Headsets, Three Failure Points
List's history with commercial headsets reads like a short catalogue of the ways consumer hardware fails. A JVC set failed at a rotating joint, which she managed to repair. A Logitech set went first at the ear sponges and then at the USB cable. The current casualty is an EPOS headset, which broke at a ball joint. In two of the three cases the fault was in the articulation between the band and the earpiece: the part that has to move to fit the ear. In her teardown, the weak point in both the EPOS and the JVC was a thin plastic stalk, one for the ball and one for the circular joint.
She considered three ways out. She could salvage a headband from one of the older dead headsets, download an existing printable design, or draw her own. She chose to design her own, and she made one decision at the start that shapes everything after it.
Why One Piece Instead of a Better Joint
The obvious approach for a maker would be to print a sturdier copy of the failed part: a two-piece joint, perhaps with thicker walls or a pin through it. List deliberately avoided that. Her reasoning, as laid out in the Hackaday piece, is that the seams where separate parts are assembled are where weak points live. Instead, she designed the headset as a single print-in-place object and relied on the elasticity of the printed material to provide the movement a joint would otherwise supply.
That is a meaningful shift in design philosophy. A conventional joint concentrates motion and load at a small contact area: a ball in a socket, a pin in a hole, a snap fit. Wear and stress pile up there. A compliant mechanism spreads that motion through a flexing section of the part itself, so there is no bearing surface to wear and no seam to separate. The trade-off is that the flexing section has to be designed carefully, and that is the job of the spring.
The Zig-Zag Spring
The earpiece mounts hang off what List calls a simple zig-zag leaf spring. According to her write-up, the geometry gives the earpiece movement in two rotational directions, X and Y, plus displacement along Z. In practical terms, that is the freedom an earcup needs to tilt and settle against the side of a head and to press in toward it, which is the work a ball joint normally does.
She built the spring as its own OpenSCAD library, kept in the repository as headset-spring.scad. Her test print of the spring by itself turned out, by her account, to be something of a fidget toy, which is a fair informal sign that a printed flexure is doing what it should. Breaking it out as a library also means the spring is not tied to the headset; anyone designing another compliant part can call it from their own code.
Inside the Repository
The repository describes the project in one line: "A print-in-place headset band intended as a replacement when your headset breaks." It is organised into a small set of files:
jennys-headset.scadassembles the parts into the finished headset.sliding-headband.scadcontains the modules for the adjustable sliding headband.headset-spring.scadholds the zig-zag spring modules.license.mdcarries the CC-BY-SA-4.0 terms.
This is not a complete headset. It is a new skeleton for the guts of an old one. The instructions say to strip the earpieces and electronics from a broken set and glue the earpieces into the printed mounts, so your donor's drivers, microphone and wiring carry over. The mount shape is the part you will almost certainly need to change. The ovoid cutout in the published file matches List's own donor headset; to fit yours, you edit the earpiece shape around line 30 of jennys-headset.scad. Because OpenSCAD is code, that adaptation means editing the cutout in the source rather than remodelling a mesh. The file's own header marks it as unfinished for exactly this reason.
It is also early days for the project. When we read it, the repository had a single commit and five stars.
What Has Not Been Tested Yet
The obvious caveat: at the time of the October 6 article, List had not yet printed the full headset. The spring has been tested as a standalone print, but nobody has reported how the assembled band performs on a head, how it holds up to months of daily use, or whether the springs fatigue faster than the joints they replace. Anyone who builds one now is effectively part of the test programme. That does not make the design less interesting; it means the durability argument for one-piece compliance is, for this headset, still a hypothesis rather than a result.
What It Means for Makers
The immediate use is plain enough. If you have a headset with good drivers and a broken hinge, this repository is a starting point for a repair that keeps the expensive parts out of the landfill, with the caveat that you will need to adapt the earpiece mount and accept an untested design.
The broader lesson is about where flexibility goes in a printed part. Many makers instinctively reproduce the architecture of the commercial object they are fixing, part for part. List's approach starts from the opposite question: which seams and joints can be eliminated entirely if the material is allowed to bend? For a print-in-place design, a flexure such as a zig-zag leaf spring can replace a hinge or a ball joint, and the result is one object instead of an assembly.
There are practical points worth taking from the structure of the project, too. Making the spring a separate, parametric library lets you print and test it in isolation before committing to a full build, which is exactly what List did. Keeping the earpiece cutout as editable source turns a one-person design into something others can adapt. And publishing under CC-BY-SA-4.0 means improvements, such as a mount for a different donor headset, can flow back under the same terms.
The real verdict will come when the full headset is printed and worn. Until then, the repository is best read as a well-organised design pattern: when the part that always breaks is a joint, consider printing a spring instead.
