AMD BC250 Case Clamshell/Horizontal Style with interchangeble front

⚠ Status: design-complete prototype, not yet fully printed. Geometry validated with PLA test pieces (tube-bend jig, PSU pattern plate, joint samples); the full case has not been printed and assembled yet. Publishing now to gather community feedback before committing ~1 kg of ABS. Printers welcome — you'd be an early tester, comments appreciated.

What it is

A horizontal, Steam-Machine-flavoured console case for the AMD BC-250 (the PS5-derived board):

• 159.6 × 477.6 mm footprint, 197.3 mm tall — sits flat under a TV

• Internal layout, front to back: front panel with power button → ATX PSU on its side right behind it (IEC facing down through the floor) → radiator + two P14 fans stacked vertically on one flank with the BC-250 vertical beside them, pump block toward the centre → a 59 mm rear bay where the AIO tubes make their return loop, next to the recessed rear I/O

• Tube routing solved for the real AIO: 30 mm bend radius, 24 mm port spacing on the end-tank, ~305/312 mm path lengths — checked numerically for interference, minimum bend radius and wall clearance. A printable R30 tube jig is included.

• 18 mm skirt/plinth all around: the IEC + rocker switch of the PSU face DOWN through an aperture in the floor (61×47 mm), a 90° C13 cable runs hidden under the case to a rear notch — zero visible cables

• USB hub bay in the skirt (rear side wall): pill slot for an Anker Ultra Slim (100×30×10 mm) resting on two internal shelves, with a Ø10 cable hole up to the board's USB — the hub's ports face outward at floor level

• Rear I/O aperture with an internal 3-wall cable tunnel (connectors recessed ~70 mm, cables exit straight)

• Dense hexagonal mesh on both flanks, roof, rear and front panel — 1620 openings, ~2.5 mm webs

The three printed parts

Two shells meeting on a lapped joint (anti-rotation notches + drilled tabs), plus a pinned front panel (4× Ø5.8 pins in Ø6 holes):

Times from OrcaSlicer at 0.12 mm / 3 walls / 20% infill on a Qidi Q1 Pro — a coarser profile will cut them drastically.

Print orientation (important!)

• Rear shell: rear face (I/O head) on the bed, joint ring up

• Front shell: front frame on the bed, joint ring up — do NOT print it joint-down, the internal frame then needs ~330 g of support towers

• Front panel: outer face on the bed, pins up

• Both shells use 237–238 mm of Z: you need a true 240 mm-Z enclosed printer (designed on a Qidi Q1 Pro, 245×245×240).

Hardware you need (BOM)

• AMD BC-250 board

• Arctic Liquid Freezer III 280 (check your tube bends with the included R30 jig first)

• ATX PSU, standard pattern (4 slotted holes in the floor; IEC faces the floor aperture — the pattern is chiral, only one orientation fits, matching the included test plate)

• PS_ON bridge/adapter for the BC-250 (e.g. mosfet.party)

• 16 mm vandal-style power button (Ø16.2)

• M3 screws + heat-set inserts for the joint tabs

• 90° C13 power cable (runs under the case)

• Optional: Anker Ultra Slim 4-port USB hub, cable ≥60 cm

Included files

• guscio_posteriore.stl, guscio_anteriore.stl, pannello_frontale.stl — this the the case
  these are tests to validate the junction ecc. with your printer 

• prov_dima_tubo.stl — R30 tube-bend jig, print first

• prov_piastra_psu.stl — PSU pattern + IEC/switch aperture test plate (61×47)

• prov_giunto_conchiglia_post.stl / prov_giunto_conchiglia_ant.stl — joint sample sections

Known open points (feedback wanted!)

1. Thermals unproven — VRAM on the board's back side is the critical spot; flank mesh is the intake, radiator side the exhaust.

2. The PS_ON adapter bay is not modeled yet.

3. The hex cells sit on ~4.6 mm internal bridges in places; fine in an enclosure, slightly rough in fast open-air PLA.

4. If you print at 0.16/0.2 instead of 0.12, report back on the joint fit.

Printer settings I use

Qidi Q1 Pro, generic ABS, 0.12 mm layers, 3 perimeters, 20% infill, normal (auto) supports, brim. Bed 100 °C, chamber closed.