Field notes
How the work actually gets made
Short, honest write-ups of the process behind a design: the measurements, the wrong turns, the numbers that decided it. Precise underneath the play.
05unprinted study
Monkey hook, inside and out
Play the fold, separate the assembly and inspect each component in the actual v0.3 CAD. Download the prototype files.
04parallel study
KIL — two ways to hang a bag
The locked outward hook and an unprinted inward candidate, at the same scale.
Spin the actual meshes, compare the mounting mechanisms, follow the chunky facets
and download the versioned models.
00printed
TANG + KLEM — the lookbook
The two-boards-one-ring family, shot as a lookbook: TANG kitchen tongs
(printed, benched, sanded, in use) and the converged KLEM clip — steel ball fulcrum and
magnet latch sealed invisibly inside.
01geometry only
Micro Tilt Display — the fold, with the hardware
A tilting 0.96" OLED desk widget on an M3 friction hinge — the first-proof hinge+wire coupon,
folding flat → open across three states + an animation, with the ESP32-S3-Zero and OLED it carries.
Collision-free across the whole fold (swept-boolean gate); unproven until the bench.
02provisional
Refitting a downloaded insert to my actual iron
A stand insert cut for the HS-02B, loose on my HS-02A. Tightening the two inner seats
without moving a point of the exterior, proven to 0.000 mm. Measure → coupon → the bug → gauge →
adversarial review.
03study
Five ways to grip a knob
The TP1 knob's reeded side, adversarially reviewed, then five printable alternatives —
each as an AI beauty shot, a CAD render, and a live 3D model you can spin. Why FDM caps how fine a knurl can go.
Mmethod
How cadgent dials in a fit
The reusable loop: the fit-gauge, the three trust rules, the tool stack, and why a render
is never a fit.