Enclosure is now fully opaque. The 355 nm window is removed, and the hinged door is replaced by a 1.5 mm alu slide-up door in 2x 2020 x 1040 tracks, balanced by counterweights. The coded interlock is on the right track, below the door.
Fully open, the door needs ~426 mm above the top panel (~1105 mm overall from the floor) and only ~60 mm in front.
Self, interface and door-travel checks: 0 hits.
All inlets are on the RIGHT panel: IEC, air bulkhead (Y 690 Z 80), pedal jack, pass-through to the external base/chiller, and the Ø100 fume port (Y 650 Z 500). The external fan is on the right.
The back and top panels are plain. Harness and hoses are rerouted (W18, W19, H04, H07; W42 pedal added).
Y-min home sensor moved under the bed carriage plate (Z 150-162) with an FDM flag; the 48-pose sweep is clean. The 2020 x 773 cross member is in the frame_motion cut list (TODO 62 done).
Focus is now MANUAL, for rare adjustment.
Removed: the focus motor, screw, coupler, bearings, knob, home switch and W15/W16.
Added: a 6 mm alu clamp bracket with 2x M5 into the column slot, a printed scale with pointer, and an FDM set stop at WD 190.
Clamp at 9 kg source: SF 4.9 static, 2.5 at 2 g, 7.4 with the stop. Top-panel clearance is unchanged (6.4 mm at WD 270).
Assumptions (pending measurement):
laser coolant/cable OD 12, R60, 1.8 m;
source 7 kg, with margins also checked at 9 kg;
pedal is a 2-wire dry contact in series with the force-guided relay.
Hot-air hose H01 is 1.64 m of 1.8 m. Laser coolant inside run is 0.28 m.
Head Z callout render added (osso_top_head_Z_callout.png).
Top regenerated: osso_top.step, combined BOM, renders, MASTER_ASSEMBLY. Sweep: 0 REAL.
Bed lift logged as a v4 idea, not adopted. Pedal-jack interlock pending the designer's wiring check (TODO 64).
2026-10-26 (later): Door shrunk to the load/unload opening (the designer)
The 1105 mm open height was too much. The front opening is now only what loading needs: 360 x 165 (X 90-450, Z 195-360). It is sized from the load pose: bed plate X 160-380 plus 70 hand room each side, the bottom 17 below the plate top, and the top at the tray top + ~75. The head is parked clear and the lens sits above the opening. The front panel around it is fixed.
New door: 390 x 195 opaque alu, lift 195, 2020 x 580 tracks on the front panel, 2x 0.22 kg counterweights. Everything stays under the top panel, so overall height stays ~680 mm and there is no overhead.
Cord-break: drop stops in the track slots mean the door can only fall to closed (laser-safe), with stops under the counterweights too.
The interlock moved with the door (right track, X 427-452), and W19 is rerouted via a grommet at X 440 Z 105.
Checks: enclosure self, interface and door travel 0. Harness travel 0 REAL. 48-pose sweep 0 REAL.
Fume: a vertical Ø100 riser now sits hard on the right wall (X 678-778, Y 390-490, Z 290-630), between the tray drive and the gantry beam. A LOW inlet bellmouth (Ø130) faces -X at Z 355, the lowest height that clears the tray bridge and flag. A Ø30 top bleed orifice is at Z 630. One tee goes to the exit (right panel Y 440 Z 560), with one external fan. The dP switch is on the right wall behind the riser, with taps in the riser rear face. The W18 and H07 routes were updated. Clearance to the head at X max / Z max is 131.8 mm. There are no clashes with the gantry, the X chain or the door.
Tray v2: 300 wide (X 107-407), sized to the station-B needle reach X 111-401. It has 8 rows x 72 pockets = 576 (was 384) and 4 pins per lane at Ø1.4 (0.1 play). Rows stay at 8 because Y is limited by the nut block vs the front fixed bearing.
No cantilever: a 2nd MGN12H is added on the LEFT rail at the tray front, with a riser and a 6 alu arm, giving 3-block support. The right-front is blocked by the tray drive (motor/coupler/bearing X 695-778, Y 319-419). Open question: move the drive?
The front opening (X 90-450) still clears the 300 wide tray. No resize.
Embossed name labels are fused onto the tops of the concept keep-out blocks (bay + pneumatic panel, blocks >= 40x25x15) in the assigned colours.
Route sweeps: the circular sweeps at assumed OD, coloured by class, were already in osso_top.step (ROUTE_*). Confirmed.
Laser mount: a second 2040 cross member (Y 345-365) and a 2020 Y rung (X 476-496) tie the column top to both cross members through a 6 alu column-top gusset. The diagonal brace and the column foot bracket are removed. The gusset lower band and the focus carriage plate top were trimmed to clear the carriage at WD 270 (1.4 mm). The rung front end is end-tapped, with no bracket.
Corner rule: the back, LEFT outer face and RIGHT guard outer face are planes, checked automatically (plane_check.json, 3 poses). The bolt-on side fan set is exempt (it marks the free side).
Real 2020/2040 T-slot profiles everywhere, including the door tracks and the X-chain shelf member.
Second (rear) MGN12 X rail, with the X carriage on both rails (bridge + rear plate). X-motor and X-min endstop cables now drop at Y 616, behind the rear rail.
Tray v3: a 3 alu carrier plus two pull-out cassettes (A rows 1-4, B rows 5-8). Each has a D4 dowel in a round hole + slot, 3 pads and a flexure latch. They are loaded on the bench with a partial cover peel. 4 pins per lane (the doc inconsistency is fixed).
Right guard on standoffs (gap 108). Lines exit top/front/bottom (WIRE/HOSE_LIST column). Hose channel for the coolant. A Velcro tie slot grid (25x5, horizontal + vertical) is in both guard sheets.
Exhaust: SIDE exit only in the model (stub → through the guard → inline fan on the guard face). The up-elbow is a docs-only option for end-of-bench use. The top stays clear.
Cart: low-profile 50 mm total-lock castors, 2020 rings, laser base on the top shelf, chiller on the lowest shelf. Lower part is open front/back/left; only the right guard runs full height. Height 1466.5 mm, bed at 1033.6 mm. Fits a wall shelf with its bottom at ≥ 58 in. Leveling feet are needed when parked.
HMI: 10.1" HDMI capacitive touchscreen on the front right (centre 1335 mm above the floor), with a Pi 5 host behind it. Button niche under it (E-stop with extended head, key, lamp, START, HOLD), ≥ 88.8 mm from the laser carriage at all WD.
Enclosure interior is matte black (powder coat, 1.5 mm 5052 kept).
Vision: head down camera + LED ring, up camera + ring light (left of the bed path, X 111), chamber overview camera, 2 diffuse LED strips. W44-W52 added.
2026-10-26 run 5 (the designer's 13 items, resume after interruptions)
Resumed: run 4/4b covered items 1-11. Open gaps found: plane check had 6 violations (side duct + bbox tolerance), the side fan sat outside the right plane, and the E-stop was 5 mm proud.
Fume fan moved INTO the cart (top shelf, right of the laser base, axis Y, outlet at the open cart front). A 80x120 rect duct runs in the guard gap (Y 520-640) from the port down into the cart. No guard hole. Laser base shifted to X 30. W21/W43 gap cables re-run at Z -25. One guard standoff moved Y 560 -> 500. Plane check has no exemptions now. Enclosure self/door/iface 0, cart hits 0.
E-stop mushroom is now 3.5 mm behind the front face (nothing proud).
Items 12-13: review-only VIZ parts (head/up/chamber camera FOV frustums, beam path source->galvo->lens, beam cone, 150x150 scan field) in cad/top/viz.py; not in the BOM or the sweeps. Review model has translucent outer panels/guards/door (BOM still opaque 5052). Exports: osso_top.step (501 named parts, RGBA colours), osso_top.glb (metres, alpha BLEND, part names), osso_top.usdz (501 named meshes, UsdPreviewSurface opacity). Plane check over 3 poses: 0 violations. Combined BOM regenerated.
48-pose sweep: 0 REAL (13 intra, 1 designed, 34 termination, 16 bundle). Site rebuilt, links checked (2 pre-existing external YouTube links flagged). No snap, no Mac sync.
Checked the run 5 items: guard Velcro strap points are pairs of parallel 10 mm obround slots (28 long, 30 c/c, H + V alternating) for a 20-25 mm strap; design notes cover the shelf fixture stops (chiller + laser base) and the E-stop alone on the niche back wall; the external panels, door and right guards export as their own 'panels' subassembly.
Top assembly rebuilt: 1103 x 773 x 1390 mm, 30 mm under a 1420 mm shelf. Bed working surface 957 mm above the floor, screen centre 1258.5 mm.
Top assembly rebuilt with the run 6 parts: the left-zone electronics plate (59 plate and pneumatic-bracket parts), the HMI niche with the reclined screen, and the tray bracket fix. 628 model parts (635 with the review VIZ items). 'panels' is still its own subassembly, separate from 'enclosure'. Harness BOM regenerated so the plate devices now appear on it. Combined BOM is 310 lines.
Overall 1103 x 773 x 1390 mm. Screen centre 1185 mm above the floor. Push-test SF on castors 1.13 front / 1.01 back / 1.56 left / 1.35 right; on feet 1.35 / 1.27 / 1.82 / 1.61.
48-pose sweep: 0 REAL (6 intra, 44 termination, 21 bundle). The tray bearing contact is gone (TODO 89 closed). Plane check now runs over all 48 poses: 0 violations at 0.5 mm, and nothing above the top panel.
STEP re-imported: 2020 sections 192.5 mm2 (33) and 2040 sections 385.1 mm2 (10), so they are T-slotted. GLB in metres with translucent panels and part names; USDZ written.
New render of the electronics plate swung out on its hook pins. Rotated 90 deg, the components face down, not up. Open question.
TODO 92-95 added: 3 tube bend-radius checks and the PU air-line radii.
Site rebuilt and links checked; not deployed. No snap, no Mac sync.
Run 7 A-C recap: exhaust inlet and front outlet at Z 540 with one elbow; fixed 676 x 558 electronics plate with the pneumatics on its bottom band; no hinge (service by removing the left panel); all hoses and wires routed inside (boundary check: 0 violations over 73 routes); hot-air heater at the top rear under the top cover, H01 hose 1.185 m.
Heat fix (conservative): the 6 printed parts within about 100 mm of the heater shield (X motor plate 53 mm, X-chain moving bracket 56, Z motor bracket 57, X nut bracket 15, X end KFL 32, Z nut block 76) are now laser-cut aluminium. A second shield with a bigger gap did not fit: the X carriage bridge is only 5 mm from the shield.
FRL bowl: only ~21 mm free below it (needs 40), so the assembly steps now say to unbolt the FRL to remove the bowl.
Top assembly rebuilt: 640 model parts (647 with the review VIZ items), 'panels' still separate. Overall 1103 x 773 x 1390 mm. Cart re-checked with the run 7 mass moves (130.3 kg, CG 649 mm): push-test SF on castors 1.14 front / 1.00 back / 1.54 left / 1.37 right; on feet 1.36 / 1.26 / 1.80 / 1.63. The back-on-castors case is right at the limit.
48-pose sweep: 0 REAL (6 intra, 42 termination, 24 bundle). 48-pose plane check at 0.5 mm: 0 violations, left face flat, top clear, height 1390.
STEP re-imported: 2020 sections 192.5 mm2 (33), 2040 sections 385.1 mm2 (10), T-slotted. GLB in metres with translucent panels and part names; USDZ written.
Site rebuilt and links checked; not deployed. No snap. Work Mac synced; home Mac could not take file copies this run (pending).
Middle accessory shelf removed: overall 993 W x 773 D x 1233.5 H mm (was 1103 x 773 x 1390); 186.5 mm clear under a 1420 mm wall shelf. Mass 120.3 kg, CG 38.8 mm above the frame bottom.
Back castors now rigid, EPDM tip pads: push SF 1.25 on castors (worst, toward the back), 1.42 on feet, no ballast (TODO 97 closed).
Right guard, standoffs and Velcro grid dropped (~110 mm narrower). All hoses and wires route inside; outside connections come through a down-facing bottom connection plate (IEC, pedal, PU8, glands). boundary_check: 68 routes, all internal.
Pneumatics (FRL, manifold, valves) moved to a plate in the top-left zone, under the top cover; the left zone is compressed vertically (the Ø160 fan sets its width). Electronics plate now 676 x 422.
Cooling: G3-filtered intake at the bottom-left of the left zone, flow bottom to top over the plate, out past the 3 cabinet fans at Z 360.
Laser base raised on a platform in the lower bay; the bundle drops at X 750-780. 1.0 m cable: 59-119 mm slack after the loop and backshells, so it fits (TODO 101 to measure).
10.1 in screen on a tilt-and-extend arm (folds flat in the niche; detents; cable inside). Standing check: stowed center 1077 mm above the floor, extended 1266; tall users (P95) look down 61° when extended, so they should stand ~150 mm back (TODO 100).
harness_checks: 0 REAL hits, mains/SELV ≥ 15.5 mm. Bend-radius items: 92 closed, 93/95 rechecked. Old swung-out plate renders replaced by service views (TODO 99).
