Transmission

Changing speed, torque or direction — gears, belts, ratchets and the parts that carry them. This shelf is built differently from the others: each entry is a parametric OpenSCAD source you can drop into a project, and the model on its page is the mesh exported from that source — the same geometry a printer would slice, not a redrawing of it.

24 of 24 rows are built. Every one of them is checked against its exported solid: no two parts may occupy the same space at any point in the motion, the faces that are supposed to carry load have to actually touch, and the number that defines the mechanism — a ratio, a travel per turn, the angle a ratchet gives back — is measured off the geometry rather than asserted.

• T1-01Spur pair

  The plain reduction between parallel shafts; the row every other gear row is measured against

  ✅ built — gate ✓ (zero intersection / 8 poses, bearing 0.186 mm), page models/spur-pair/

• T1-02Internal / ring gear

  An internal mesh turns the same way as its pinion and packs a big ratio into a small circle — the ring of a planetary

  ✅ built — gate ✓ first try, bearing 0.180 mm; vision LOOK (the pinion needs a visible mark to prove same-direction)

• T1-03Idler

  A third wheel that reverses direction and changes nothing else; the honest way to show that a ratio is set by the ends, not the middle

  ✅ built — gate ✓ (zero intersection / 48 poses, bearing 0.191 mm); the idler's tooth count is not in the ratio

• T1-04Rack and pinion

  Rotation → a straight line, at exactly π·m·N per revolution. The relationship is the row

  ✅ built — gate ✓, and the travel came back 100.53 mm = π·2·16 exactly

• T1-05Compound train

  Two pairs on one shaft: ratios multiply. The row that makes a seed meter's odd ratio reachable

  ✅ built — gate ✓ first try, both meshes bearing; 2.5 × 2.0 = 5:1, output 72° per input turn

• T1-06⭐ Planetary set

  Sun, planets, carrier, ring — a coaxial reduction in one stage. The robot-joint row

  ✅ .scad built, gate ✓ first try — 6 parts, all six meshes bearing, ring stays, planets orbit 41.9 mm. The phasing is the hard part and it landed

• T1-07Cycloidal drive

  A lobed disc on an eccentric: a very high single-stage ratio, and printable where a planetary is not

  ✅ built — derived epitrochoid, 10:1 opposite; gate ✓ 121 poses, disc on pins 0.198 mm

• T1-08Bevel pair

  Power round a corner, 90° between shafts

  ✅ .scad built, gate ✓ first try — bearing 0.122 mm

• T1-09Worm and wheel

  A large ratio in one step that will not back-drive — the row a lifting mechanism wants

  ✅ built — single start, 1:30; gate ✓ 121 poses, bearing 0.041 mm; prints standing, lower flanks droop (accepted)

• T1-10Helical pair

  Quieter than spur, at the cost of an axial thrust the bearing has to take

  ✅ built — gate ✓ 48 poses, bearing 0.171 mm; thrust coefficient 0.364

• T1-11Herringbone pair

  Helical's thrust cancelled by mirroring it — and it prints without support

  ✅ built — gate ✓ 61 poses, bearing 0.159 mm; prints without support

• T1-12GT2 pulley

  The part the drawing robot lists as needed: 2 mm pitch, 20T, on the recorded NEMA 17 D-shaft

  ✅ built — gate ✓, belt bearing 0.100 mm; page LOOK (one oblique view cannot show tooth phase)

• T1-13Belt tensioner

  An idler on a slot; the thing that makes a belt drive work and is always left out. The drawing robot lists one

  ✅ built — gate ✓ first try; 2 mm take-up per mm of slot, 20 mm over the 10 mm range

• T1-14Roller-chain sprocket

  #25 · #35 · #40, the farm size

  ✅ built — #40, 15T, rollers seated at 0.000 mm (48 poses); ANSI B29.1 form, compliance unverified

• T1-15Face wheel and lantern (cog and rung)

  A wheel whose cogs stand out of its face drives a lantern of round staves on an axis crossing it at 90°: the oldest right-angle drive, and the whole train of a colonial gristmill (Betz's Mill, HAER PA-64: 60 cogs → 24 rounds, 54 → 13)

  ✅ built 2026-10-17 — 30:12 at p 10 mm; gate ✓ 181 poses over 3 pitches, no overlap; running clearance 0.20–0.78 mm (nominal 1.25), tightest 0.197 on the leading flank at the cog's outer end 8° past centre — the rest pose; gate 7 ✓ (the second disc is a cap on the rounds' tenons); sees PASS; section cut from the solid · for Betz's Mill (../../assemblies/betz-mill/)

• T1-16Great spur wheel and two stone nuts

  One big wheel on an upright shaft drives a small one on each side of it, so one water wheel turns two pairs of millstones, each faster: the English mill's train above the wallower (Peirce Mill, HABS DC-22 sheet 3: 85 cogs → two nuts of 19, 4.47 : 1, the nuts 146.5° apart)

  ✅ built 2026-10-17 — 45 : 10 at module 1.5, the nuts 144° apart (18 wheel pitches, so both mesh in one phase, as the assembly sets them); gate ✓ 97 poses, no intersection, both meshes bearing 0.190 mm; gate 7 ✓ (prints flat, no supports); page probe, shape, works, bears ✓; sees PASS · for Peirce Mill (../../assemblies/peirce-mill/)

• T1-20One-way / sprag clutch

  Drive one way, free the other

  ✅ built — split outer race printed face-down, 0.8 mm leaf springs; gate ✓ 181 poses, every roller and spring 0.100 mm

• T1-21Ratchet and pawl

  The seeder's row: backing the ground wheel up must not meter seed

  ✅ built, 2 rounds — gate ✓ both directions; lock_deg = 21.01°, and in reverse the pawl stays.

• T1-22Geneva wheel

  Continuous rotation → indexed steps; a seed plate that stops at each cell

  ✅ .scad built — and proved as a Geneva: the cross is held 0.00 mm for 75 % of the turn and indexes 8.62 mm in its window. It declared motion_windows before the gate could read them

• T1-23Printed torque limiter

  A stone in the seed plate must not strip it — a detent that slips at a set preload

  ✅ built — offset-stroke fingers (the round-joint slivers fixed); gate ✓ 121 slipping + 121 driving, pads 0.100 mm

• T1-24Lantern slid out of mesh by a sliding bearing block

  One wheel, two pairs of stones, one at a time: the lantern's shaft runs in a wooden block that slides in a trench in the spur block, along the master wheel's axis, so the rounds come straight out from between the face cogs; a wedge-shaped locking block dropped in behind holds it in mesh, and a lever on a fulcrum pin slides it; done only with the wheel stopped (HAER PA-64 sheet 6, details A/B)

  ✅ built 2026-10-17, reworked the same night (round 1 lifted the block; the record slides it) — 9 solids (5 mm shaft and M3 pin bought); gate ✓ 121 poses, no overlap, 10 bearing pairs; motion_windows: the wedge out first (0–0.3), then a 7.5 mm slide (0.3–1), the rounds a declared 3.0 mm past the cog tips; lever 3.75:1; gate 7 ✓ (teardrop pin hole); sees PASS; sections cut from the solid · for Betz's Mill (../../assemblies/betz-mill/)

• T1-30Four-bar linkage

  The crank-rocker, and the row that says a linkage is a ratio too

  ✅ built — Grashof crank-rocker; rocker swings 45°, minimum transmission angle 53.8°

• T1-31Crank-slider

  Rotation → reciprocation, with the stroke as the parameter

  ✅ built — gate ✓ 145 poses, all joints 0.100 mm; the slider is not at mid-stroke at 90° of crank

• T1-32Disc cam + follower

  An arbitrary motion profile from one turn — the general case the others are special cases of

  ✅ built — cycloidal rise/dwell/return/dwell; follower held 0.00 mm in both dwells; max pressure angle 15.4°