# linear_cam_short_v4: v3 head + rev B nozzle stack (head_linear_cam_v4)

Built from v3 (v3 unchanged); approved review: drawings/head_linear_cam_review_r2 (rev B, "Ready for CAD" 2026-10-29). Scripts: model.py, checks.py, sweep.py, clearance.py, export.py. mm throughout.

## v4 change (only this; all other v3 open items left as they were)
- All 6 stations: standard 304 SS tube 4.0 x 0.5 x 76.2 bonded 5 into a D4 socket in the slider bottom; ISO 80369-7 male Luer-lock adapter (6 socket, 11.7 body, hex AF8); equal all-metal 1/2 in Luer-lock tips (Ni-plated brass hub 11.5 long x Ø6.0 body, Ø7.8 lugs — hub dims are ESTIMATES, TBD, no vendor drawing found; 304 SS shaft 12.7 exposed), 25 ga N2..N6, 23 ga N1. Replaces the v3 Luer hub + 2 in / 2.5 in (N1) needles.
- Base plate: CUSTOM D7 bore per station (was D3). CUSTOM guide bushing (option C): OD7 body Z -10..6 through the bore, flange D10 x 1.5 on the base underside (Z -1.5..0); ID custom-fit to the tube stock (4.05 nominal modelled). Deviation from the r2 drawing: flange drawn at Z -11.5..-10 there (not touching the base); modelled against the base so the bushing is grounded.
- Stack below the slider boss 101.6. Tips (local Z, base bottom = 0): all six -51.6 at HOME / -81.6 active (v3: N2..N6 -12.8/-42.8, N1 -25.5).

## v4-N1 change (pt1194u, 2026-10-29): N1 = BALL DROP ONLY (never picks, gravity drop, no air puff)
- New N1-only mSLA body `custom/slider_mSLA_tower_N1`: v4 slider + integral tower (Ø3.6 strut), Ø0.40 ball channel at **31°** from axial, azimuth **27°** (from -Y toward -X), 54.7 long, merging into a Ø3.0 bore (Z 55-57.6) on the axis at **Z 56**; funnel at the inlet under the wheel pocket exit. Capped blind port boss (Ø4 x 3, no hole) on the tower for a future puff. No cross-port.
- Angle/azimuth changed from the rev A section (28° / 25.6°): at 28° the channel is inside the rail carriage (Y ±20) at Z 98, and the collar gap was 0.43. Section drawing refreshed to rev B.
- r3 feeder (housing, lid, guide ring, wheel, gearmotor, IR sensor, fill plug, OPTIONAL vibe) moved rigidly on N1 by X -9.19 / Y -7.50 / Z +32.88 (inlet (-12.79, -25.10, 102.88)). Separate feeder chute deleted (the channel is in the tower).
- Removed: N1 PC4 elbow, N1 PU 4x2.5 hose, 2/2 puff valve, feeder PU2 air line (the PU2 clash with cam_hanger is gone with it). Bundle now wires only.
- N1 tip 23 ga (OD 0.64 / ID 0.33); N2..N6 25 ga; all tips equal length (-51.6 HOME / -81.6 active).
- N1 stop collars Ø5 (others Ø6) + local relief cut in the tower: tower-to-collar 1.05 (was 0.43), channel-to-collar 2.07.
- Bundle moving leg follows the housing (X -33.2, Y -27.5); loop length constant 93.5 (was 60.6), R 8.
- Margins (n1_margins.py / _n1_margins.json, BRep, sweep): tower to rail carriage 0.53 (corner, trimmed face Y -20.9 above Z 97.5), cam_hanger 0.90, cam plates 0.70 (same as the fork), slide body 5.5, roller 0.44, axle 0.94; min channel wall 0.90 (tower), 0.27 at the fork root Z 67.7 (inside the body, slot side).
- Mass: N1 moving 39.0 g (v4 57.2 g, incl. half the bundle loop; N2..N6 21.9 g). Spring: with the standard spring SF ~1.96 (3.5 x 21.9 / 39.0); fit an N1 spring with ~1.8x the force to keep SF 3.5. Head total 2.229 kg (v4 2.249), CG (49.0, 12.5, 67.3).
- Envelope unchanged (bbox HOME and active identical), so run15 tray sweep not re-run.

## Envelope and mass
- 289 L x 148.5 W x 201.6 H at HOME (Z -51.6..150; v3 175.5 H); 231.6 H with a station active (Z -81.6).
- 2.249 kg (v3 2.197; +52 g); nozzle stacks x6 incl. bushings 64.1 g (adapter as solid SS by volume: conservative). CG local (48.5, 12.3, 66.9).

## Checks
- checks.py (HOME + N1..N6): 0 clashes, 0 floating (148 parts). sweep.py (84 poses, full carriage travel): 0 clashes.
- clearance.py (_clearance.json): bushing-to-adapter 5.2 at HOME, 35.2 active; bushing-to-slider 44.0 HOME, 14.0 active; tube radial in bushing 0.025 (nominal ID); tube-to-base bore 1.5 radial; adapter-to-neighbour adapter 12.0 (12.05-12.49 while moving); active tip 20.75 below the lowest neighbour (neighbours drop 9.1-9.25 on the v3 cam ramps, as v3).
- N1 (v4-N1): ball path resolved by the tower channel into the Ø3.0 bore above the tube (see v4-N1 change).

## Tray / head (run 15, integration/subsystems/top/run15/summary.json)
- Active board-touch Z 355.2 = Z_MIN_SOFT (was 316.4). Needle-down only with tray T >= 462. Z_IL 361 at T_EXT, Z_IL_BACK 372 at T_BACK. X soft 160..425 unchanged.

## Exports (repo cad/, flat)
- head_linear_cam_v4.step (HOME; top-level components / custom / panels), head_linear_cam_v4.glb (front view, panels translucent), head_linear_cam_v4.usdz (static Quick Look, HOME). No animation. Preview: drawings/2026_10_29_head_linear_cam_v4_preview.png.

## What changed from v2
- r3 feeder: FR4 3-layer pocket wheel, guide ring, chute 59.2 deg x 33.8 mm through the N1 slider channel, IR sensor, 2/2 puff valve and PU2 air line (puff lands the ball in the flux).
- OPTIONAL coin vibration motor on the feeder lid.
- Enclosure: 5 x 1.5 alu laser-cut panels (hidden in animations); only needles protrude below; top exit with brush grommet for hoses and cables.
- Cheap fixes: chamfered cam ramps, mSLA slider forks, slotted motor face plate (belt tension), FDM belt cover, flex-loop feeder bundle.
- Gantry mount: 2 ribs + adapter plate. Run 10: down camera housed inside the head (front-middle pocket) looking through a D30 base window.

## v3 checks still valid
- Hose length constant 163.6 mm, min bend R 11. Bundle length constant 60.6 mm, R 8 (R/D 1.6 - see open issues).
- Spring return SF ~3.5.

## v3 exports (in cad/versions/head_linear_cam_short_v3)
- STEP per pose (HOME, N1..N6 active, N1park20) with components and panels subassemblies.
- GLB per pose + home_enclosed.glb, all opening on the front view.
- linear_cam_short_v3_anim.usdz (Quick Look, 277 frames, 12 fps); linear_cam_short_v3_anim.mp4 (23.1 s, front view, feeder inset).

## BOM
BOM_linear_cam_short_v4.csv (no prices).

## Open issues
- v4-N1: N1 spring needs ~1.8x force (SF 1.96 with the standard spring) - spring part TBD.
- v4-N1: wheel pockets / guide ring are still the r3 Ø0.5-ball design; resize for Ø0.25 balls.
- v4-N1: tower cantilever stiffness (feeder ~20 g on a Ø3.6 resin strut, plus the housing floor joint) not analysed; add a rib if it flexes.
- v4-N1: 0.53 tower-to-carriage at the trimmed corner; 0.27 channel wall at the fork root (1 mm long) - check mSLA print.
- NEW v4: bushing ID / material TBD (custom-fit to the measured tube).
- NEW v4: tray rule much tighter: needle-down needs T >= 462 because the stack is 101.6 deep; rp bracket / upcam at x 425 for T 432..452.
- Feeder bundle dynamic bend R/D 1.6 is below the usual 5-10x for continuous flexing: use a high-flex silicone wire bundle or lengthen the loop.
- No hot-air nozzle station in this head (v2 head had station E). Per the design note, the hot-air station will also use a metal Luer needle when it comes back; none is modelled.

## Design note: all-metal Luer tips (pt1198u, 2026-10-29)
- All stations use all-metal Luer-lock tips (metal hub + 304 SS shaft), 1/2 in standard length. Plastic hubs are not used because of reflow heat. The hot-air reflow station also uses a Luer needle so its gauge can be changed if testing shows a gauge change helps.
- Result (rebuild): tip Z -51.1 HOME / -81.1 active (was -51.6 / -81.6; metal hub 11.5 est. is shorter than the PP hub 18.35, so the 1/2 in tip ends 0.5 higher). Boss -> tip 101.1.
- Checks: 0 clashes / 0 floating at HOME, each station active, 84-pose sweep. Mass 2238 g (+9 g; nozzle stacks x6 73.0 g); N1 moving 40.4 g, spring SF ~2.0.
- Run 15 re-run (top/run15/metal_tips/summary.json): needle-down still T >= 462 (min gap 2.0, N4 adapter); Z_BOARD / Z_MIN_SOFT 354.7; Z_IL 361 and Z_IL_BACK 372 kept. Needle hits now checked at 1e-3 mm3.

## Approved (the designer, 2026-10-29)
- Hot-air (E) nozzle placement APPROVED: it stays on the head-front bracket (as in TOP run 16).
- E-only Z soft limit APPROVED: Z_MODE_E = 327.7, allowed only when T >= 462 (tray clear). All other stations keep Z_MIN_SOFT 354.7.
