A Chinese Windlass is a hoisting mechanism built around a single shaft with two drums of slightly different diameters, where a single rope winds onto the larger drum while simultaneously unwinding from the smaller one, with the load hung in the bight. Unlike a simple windlass, which gives a fixed mechanical advantage based only on crank length, this design multiplies advantage by the ratio of the drum diameters, letting you lift several hundred pounds with a few pounds of hand force at the cost of slow lift speed.
Chinese Windlass Interactive Calculator
Vary crank length, drum diameters, and load to see mechanical advantage, hand force, lift per turn, and torque.
Equation Used
The Chinese windlass gains advantage from the small difference between the two drum diameters. For each crank revolution, the load rises by half of D1 minus D2, while the ideal mechanical advantage is 2L divided by that same diameter difference.
- Ideal rope and bearings; friction losses ignored.
- Large drum diameter must be greater than small drum diameter.
- Single-layer rope winding keeps effective drum diameters constant.
How the Chinese Windlass Works
The Chinese Windlass, also called the Chinese Wheel in some old rigging texts, runs on one principle: every full turn of the shaft, the rope gains length on the big drum and loses length on the small drum, but only loses a smaller amount. The load — hung from a pulley block in the loop between the two drums — rises by half the difference per revolution. Make the two diameters close together and you get massive mechanical advantage. Make them far apart and you get a fast lift with less advantage. That ratio is the entire design knob.
Why build it this way at all? Because a normal drum hoist needs either a long crank or a gear reduction to lift a heavy load by hand. The Chinese windlass (compound differential) bakes the reduction into the drum geometry itself — no gears, no chain, no second shaft. One piece of timber or steel, turned to two diameters, with a hand crank on the end. That is the whole machine.
Tolerances matter more than people expect. If the rope wraps overlap on the larger drum because you cut the helical groove too shallow, the effective diameter grows mid-lift and your mechanical advantage drifts. Same problem in reverse on the small drum. If the bearings on the shaft drag — common with a wood-on-wood pillow block that has dried out — you can lose 20 to 30% of input torque to friction before the load even moves. And if the rope stretches differently on the two drums (one side takes load, the other does not), the load can drift sideways or the bight can foul. Builders working from the original Chinese windlass (form) shown in 19th-century mechanics textbooks usually cut a clean spiral groove on each drum at one rope diameter pitch to keep the wraps single-layer and predictable.
Key Components
- Large drum (D₁): The drum that takes up rope as the shaft rotates. Diameter is typically 100 to 200 mm on a hand-cranked unit. Cut a helical groove at one rope-diameter pitch — overlap on the wraps changes effective D₁ and shifts the mechanical advantage during the lift.
- Small drum (D₂): The drum that pays out rope on the same shaft. Diameter is usually 80 to 90% of D₁ — that gap sets the entire mechanical advantage. A 10 mm difference between two drums of 100 mm and 90 mm gives roughly 20:1 advantage at the load block.
- Single shaft: Both drums turn together as one rigid piece. Run-out between the two drum surfaces should stay under 0.5 mm or the rope feeds inconsistently, especially on long lifts. On a wood prototype, turn both drums in one setup on the lathe — do not glue them on separately.
- Pulley block (load block): The block hangs in the bight of the rope between the two drums and carries the load. A standard sheave with a 25 to 40 mm bore rides comfortably on a 12 mm pin for loads up to about 500 lbs. The block must spin freely or the rope chafes and the lift judders.
- Hand crank: Provides input torque. Crank radius typically 200 to 400 mm — longer arms multiply the already-large advantage but slow the lift to a crawl. A 300 mm crank on a 20:1 drum ratio means 60:1 overall, which is why old well-builders could lift a full bucket of stone with two fingers.
- Rope: Single continuous length, anchored at one end to the large drum and at the other to the small drum, passing through the load block. Manila or polyester at 8 to 12 mm diameter. Rope stretch under load matters here — low-stretch double-braid polyester gives a much steadier lift than three-strand manila on heavy loads.
Where the Chinese Windlass Is Used
The Chinese Windlass shows up wherever you need a big lift force from a small hand input and you can tolerate a slow lift. Wells, theatre rigging, mine shafts, ship deck gear, classroom physics demos — the same compound differential drum geometry, just scaled. The differential windlass term is more common in modern engineering texts; the Chinese windlass (compound differential) name sticks in historical and rigging contexts.
- Water wells and rural construction: Traditional stone-lined wells across northern China and parts of central Europe used a wooden Chinese windlass to lift full buckets of water or excavated soil from depths of 10 to 30 m using a single operator. The mechanism appears in Agostino Ramelli's 1588 'Le diverse et artificiose machine'.
- Theatre and stage rigging: Pre-electric theatres including the Drottningholm Court Theatre in Sweden used differential drum hoists derived from the Chinese Wheel principle to fly scenery flats and chandeliers. Operators could hold a 200 lb chandelier in mid-air with one hand because the drum ratio gave 30:1 advantage.
- Educational demonstration: MIT's mechanical engineering teaching collection and the Smithsonian's mechanical models from the Coolidge collection both include working Chinese windlass models used to demonstrate compound mechanical advantage to first-year students.
- Mining and shaft work: Cornish tin mines and Welsh slate quarries used hand-operated differential windlasses for surface-level material handling on shafts under 15 m before steam winches took over in the late 1800s.
- Marine deck equipment: Small fishing vessels in coastal Fujian and parts of the Mediterranean historically rigged differential drum windlasses for hauling nets and small anchors, where the slow lift is acceptable and the high force lets a single crewman do work that would otherwise need three.
- Hobby and maker projects: Modern wood-shop builders use the Chinese windlass form for shop-built engine hoists, garage door counterweights, and small sailboat trailer winches, often built from two turned hardwood drums on a single steel shaft.
The Formula Behind the Chinese Windlass
The mechanical advantage of a Chinese Windlass depends only on three numbers: the crank radius, the average drum radius, and the difference between the two drum diameters. The smaller you make that difference, the higher the advantage — but the slower the lift, because each turn of the crank moves the load only by half the diameter difference times π. At the low end of the practical range, a 5 mm difference between drums gives huge advantage but the load creeps up at maybe 8 mm per crank revolution. At the high end, a 30 mm difference lifts noticeably faster per turn but you lose much of the force advantage. Sweet spot for most hand-built units sits at a 10 to 15 mm difference on drums of 100 mm nominal — that's where you get a useful 30 to 50:1 overall advantage and a lift speed you can live with.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| MA | Mechanical advantage — ratio of load lifted to hand force applied at the crank | dimensionless | dimensionless |
| Rcrank | Crank handle radius from shaft centreline | m | in |
| R1 | Radius of the larger drum (rope take-up side) | m | in |
| R2 | Radius of the smaller drum (rope pay-out side) | m | in |
| vload | Load lift speed per crank revolution | m/rev | in/rev |
Worked Example: Chinese Windlass in a workshop-built engine-stand hoist
You are building a Chinese windlass from two turned hardwood drums to lift a 400 lb wood-stove cast-iron flue assembly up to a second-storey installation point in a heritage barn renovation. The big drum measures 110 mm in diameter, the small drum 95 mm, and the hand crank arm is 300 mm. You want to know what hand force the lift will need at nominal geometry, and how the numbers shift if you re-cut the small drum to 100 mm (closer match) or 85 mm (wider gap).
Given
- D1 = 110 mm
- D2 = 95 mm
- Rcrank = 300 mm
- Load (W) = 400 lbs
Solution
Step 1 — at nominal geometry, compute the difference in drum radii:
Step 2 — apply the mechanical advantage formula at nominal:
Step 3 — solve for hand force needed at the nominal 400 lb load:
That's a featherweight pull. To put it in context, lift speed at nominal is π × (D1 − D2) / 2 = π × 15 / 2 ≈ 23.6 mm per crank turn — so to raise the flue by 3 m you crank about 127 revolutions. Slow but trivial in effort.
Step 4 — at the low end of the practical range, re-cut the small drum to 100 mm (a 10 mm difference becomes 5 mm radius gap, which is a tighter match):
Now the load lifts at only ≈ 15.7 mm per revolution — you'd need 191 turns for the same 3 m lift. The hand force is comically low but you'll be cranking forever, and any rope-overlap error on the drums shows up as a much bigger MA wobble because the difference is so small.
Step 5 — at the high end, cut the small drum to 85 mm (wider gap):
Now lift speed jumps to ≈ 39.3 mm per turn — only 76 cranks for 3 m — but you're pulling harder. Still trivial for one person but you feel it now.
Result
At nominal geometry the hand force needed is 5. 0 lbs to lift a 400 lb load — light enough to do one-handed for hours. The low-end build (5 mm radius gap) drops that to 3.3 lbs but doubles the cranking time, while the high-end build (12.5 mm gap) raises hand force to 8.3 lbs but cuts crank count nearly in half. The 7.5 to 10 mm radius gap range is the practical sweet spot. If you measure 10 lbs of hand force instead of the predicted 5 lbs, check three things: (1) shaft bearing drag — a dry wooden pillow block can eat 30% of your input torque and the symptom is a load that won't start moving until you really lean on the crank, (2) rope overlap on the large drum where the wraps have stacked into a second layer, which raises effective D₁ and cuts MA mid-lift, and (3) load-block sheave seizure — a sheave that won't spin freely turns the whole rig from a 2-fall system into a single rope, halving your advantage instantly.
When to Use a Chinese Windlass and When Not To
The Chinese Windlass solves one specific problem — huge force advantage from a single shaft with no gears — but it pays for that with lift speed and rope handling complexity. Compare it against the two mechanisms a builder normally weighs against it: the simple windlass and the block-and-tackle. The differential windlass and the Chinese windlass (form) are the same machine under different names, so they share a single column here.
| Property | Chinese Windlass | Simple Windlass | Block and Tackle |
|---|---|---|---|
| Mechanical advantage range | 20:1 to 200:1 typical | 3:1 to 8:1 typical | 2:1 to 16:1 typical |
| Lift speed (per input revolution or pull) | 10 to 40 mm/turn — slow | 200 to 600 mm/turn — fast | depends on hand pull length, moderate |
| Load capacity (hand-built unit) | up to ~1000 lbs | up to ~300 lbs | up to ~600 lbs |
| Build complexity | Two precise drum diameters on one shaft, helical groove | Single drum, trivial to build | Multiple sheaves, blocks, and rope reeving |
| Rope length consumed per metre of lift | ~130 m of rope per 1 m lift at 100:1 MA — huge | 1 m rope per 1 m lift | MA × 1 m of rope per 1 m lift |
| Sensitivity to manufacturing tolerance | Very high — 1 mm error on D₂ shifts MA by 10%+ | Very low | Low to moderate |
| Best application fit | Slow heavy lifts where speed doesn't matter (wells, scenery) | Routine moderate-load hauling (boat trailers, anchor) | Variable MA needs and overhead lifts |
Frequently Asked Questions About Chinese Windlass
Because the rope wraps are stacking into a second layer on one of the drums, which changes the effective drum diameter mid-lift. As soon as the large drum gets a second-layer wrap, its effective D₁ goes up, but D₂ on the small drum hasn't changed yet — so the gap (D₁ − D₂) widens and your MA drops. You'll feel it as a sudden increase in crank force halfway up.
Fix it by cutting a single-layer helical groove at exactly one rope-diameter pitch on both drums, and limit the lift height so the rope never runs out of groove. For a 100 mm drum and 10 mm rope, a 30-turn groove gives roughly 3 m of single-layer wrap — that's your usable lift range.
Yes — same mechanism, different name. 'Chinese Wheel' shows up in 19th-century English rigging and theatre texts, while 'Chinese windlass (compound differential)' is the term you'll see in modern mechanical engineering references. The compound differential drum on a single shaft is the defining feature in both names.
Decide on three axes: how often you'll use it, how fast you need the lift, and how much overhead headroom you have. A Chinese windlass gives you 50:1 to 100:1 advantage in a compact one-shaft package — great for permanently installed lifts where you crank once a week and don't care that it's slow. A block-and-tackle gives you faster lift but typically caps at 8:1 to 10:1 in practical hand-rigged form, and it eats overhead height because the blocks need separation.
Rule of thumb: lifts under 2 m where speed matters → block and tackle. Lifts over 3 m where you want to do it one-handed and don't care about speed → Chinese windlass.
Two usual causes. First, the load block sheave is stuck — if it can't rotate freely, the rope on one side carries all the load and the other side goes slack. Test by spinning the sheave by hand with no load; it should free-spin for at least 2 seconds.
Second, the two drums have different rope grip. If you used new polished rope on a smooth drum surface, the rope can slip on the take-up drum while still feeding correctly on the pay-out drum. Roughen the drum surface or switch to a slightly rougher rope construction (three-strand instead of double-braid) to give it bite.
You can, but think carefully about the input speed. A Chinese windlass already has huge mechanical reduction built in — adding a 1750 RPM motor on top means lift speeds in the 0.5 to 2 m/s range with massive torque, which is dangerous on a hand-built unit. The drum bearings, the rope, and the load block were never designed for that.
If you must motorize, use a gearmotor in the 20 to 60 RPM output range and add a slip clutch or torque limiter on the input shaft. Otherwise the first time the load fouls something, the motor stalls and the rope snaps or the wood drum splits along the grain.
A reasonable rule is rope diameter ≤ D2 / 10. So for a 95 mm small drum, use rope no larger than 9.5 mm — 8 mm is a good practical choice. If the rope is too thick relative to the drum, the bend radius fatigues the fibres on every cycle and you'll see strands fail within a few dozen lifts.
Also remember that rope diameter sets your groove pitch. A 9 mm rope on a 1.0 mm-pitch error groove will cross-wrap within 10 turns. Match the lathe groove cut to the actual rope diameter, not the nominal — measure your rope under tension first.
References & Further Reading
- Wikipedia contributors. Differential windlass. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
