DIY 3D Printed Fight Stick Guide: Load Paths, Materials and CNC Upgrades

• 1
  Where Force Actually Goes in a DIY 3D Printed Fight Stick
• 2
  Choosing Material and Wall Thickness for the Case
• 3
  Print Parameters That Decide Whether the Case Survives
• 4
  When a Machined Bracket Beats Any Print
• 5
  Step by Step: From CAD to a Working Enclosure
• 6
  Plastic or Metal: Which Fight Stick Parts to Print and Which to Machine

A fight stick is not a static box. Every quarter-circle on the lever feeds a side load into the mounting plate, and that load has to reach the desk through the case walls. On a printed enclosure the load path usually runs through layer lines rather than solid material, which is why failures cluster at the same four places. Print orientation decides whether a wall is strong or weak.

The lever mount takes the worst of it. A Sanwa-style JLF lever sits on a 90 mm × 90 mm plate pattern, and a hard flick can push 30–50 N sideways through two or three M4 screws. If those screws thread straight into plastic, the threads creep within a season of play. Heat-set brass inserts spread that load over a larger cylindrical face and hold torque far better.

The button plate sees a different load. A 30 mm button is pressed with 1–2 N, which is trivial. What matters is deflection: a plate that flexes more than about 0.5 mm under thumb pressure changes the feel and eventually cracks at the snap-fit tabs. Rib the underside instead of thickening the whole plate.

Case corners and the seam between the top and bottom shell carry impact. A stick that slides off a lap and lands on a corner loads a small area with a short, sharp impulse. Rounded corners, 3–4 mm wall thickness and a gasket seam handle this better than a thin shell with sharp edges.

• 1
  Layer direction mattersPrint lever bosses so layers stack along the screw axis, not across it.
• 2
  Inserts over tapped plasticHeat-set M4 inserts for any fastener you expect to remove more than twice.
• 3
  Ribs beat thicknessA 2 mm rib adds stiffness with less mass than a 5 mm wall.
• 4
  Rounded cornersA 3 mm radius spreads impact instead of concentrating it.

For a DIY 3D printed fight stick, PLA is the easiest starting point and the worst long-term choice. It is stiff, prints cleanly and costs little, but it creeps under sustained screw preload and softens in a hot car. A stick left in a closed vehicle at 50 °C will relax its own bolt bosses. If the enclosure never leaves an air-conditioned room, PLA is still fine.

PETG is the pragmatic middle. It takes 0.2–0.3 mm more wall for the same stiffness, but it tolerates impact and does not creep the way PLA does. Print at 240–250 °C with a 0.4 mm nozzle, 0.2 mm layers and 4 perimeters. Four perimeters at 0.45 mm extrusion width gives a 1.8 mm shell that behaves like solid material in the corners.

ABS and ASA are what production arcade housings use. ASA holds up to sunlight and handles a drop without shattering, but it needs an enclosure to print without warping. For a one-off build, the extra setup rarely pays off unless the stick travels.

The real jump is metal. A machined aluminum lever bracket or bottom plate removes the creep problem entirely. Aluminum 6061-T6 is the usual pick: it machines fast, anodizes cleanly and weighs about a third of steel. Carbon fiber and PC blends sit between plastic and metal, but they cost more than aluminum and still creep under preload.

• 1
  PLAStiff and cheap. Indoor builds only. Creeps under screw load.
• 2
  PETGBest all-round filament for a first build. 4 perimeters, 0.2 mm layers.
• 3
  ABS or ASATravels well, needs an enclosure, tolerates heat and drops.
• 4
  Aluminum 6061-T6For lever brackets and base plates. No creep, anodizes cleanly.

Wall count is the single setting that matters most. Three perimeters is a display piece. Five perimeters with a 6 % gyroid infill is a case you can lean on. Infill percentage is far less important than the shell, because bending stress lives in the outer skin. Set top and bottom layers to at least five so the button plate does not oil-can between ribs.

Hole size is the second trap. FDM holes print undersize by roughly 0.1–0.2 mm because the nozzle path cuts the corner. A 24 mm button hole drawn at 24.0 mm typically measures 23.8 mm. Draw button openings at 24.2–24.4 mm and test-fit one before printing the second half of the shell. The same rule applies to 30 mm buttons.

Threaded holes should not be printed as threads. Print a 4.2 mm pilot for an M5 heat-set insert, and let the insert define the thread. If you must tap directly into plastic, use a coarse thread, keep engagement under 1.5× diameter, and expect to re-tap after a few removals.

Layer orientation on the lever plate is worth planning before you slice. If the plate prints flat, the screw bosses are loaded across the layer plane and split along layer lines. Rotate the part or add a separate boss geometry so the pull-out force runs along the extrusion direction.

• 1
  Perimeters4–5 walls at 0.45 mm width. Shell carries the bending.
• 2
  Button holesDraw 0.2–0.4 mm oversize, then test-fit one button.
• 3
  Inserts4.2 mm pilot for M5 heat-set inserts, installed at 240 °C.
• 4
  Top and bottom layersFive or more, to stop the plate from oil-canning.

There is a point where more plastic stops helping. A lever bracket printed in PETG at 5 perimeters will still creep at the screw bosses after a few hundred hours of play, because the polymer relaxes under constant preload. Adding ribs raises stiffness but does nothing for creep. That is a material limit, not a design failure.

A machined aluminum bracket solves it in one move. M4 threads tapped into 6061-T6 hold torque indefinitely, and the part stays flat. Flatness matters more than people expect: a lever plate that is 0.2 mm out of plane tilts the gate and makes diagonals feel inconsistent, even though the electronics are identical.

The same logic applies to the base plate. A 3–4 mm aluminum base adds mass low in the case, which keeps the stick from sliding during hard inputs. Printed bases work, but they need weight pockets filled with steel or they skate on a smooth desk.

For low-volume builds, CNC is more accessible than it sounds. No minimum order quantity means one bracket is a valid order, and aluminum 6061 machines quickly. Anodizing in clear, black or a color keeps the part looking like a finished product rather than a prototype. If you want to keep the printed shell, send only the two or three parts that carry load.

• 1
  Creep is the limitPolymer bosses relax under preload; metal threads do not.
• 2
  Flatness drives feel0.2 mm out of plane tilts the gate and hurts diagonals.
• 3
  Mass low, not highA dense base plate stops the case from sliding.
• 4
  One part is fineNo minimum order quantity, so a single bracket can be machined.