A Compound Weight Motor is a gravity-driven prime mover that converts the potential energy of one or more falling weights into rotary or linear work through a compound pulley or gear train. A typical tower-clock weight drive delivers 1 to 5 W of steady output for runtimes of 8 days on a single wind, with torque variation under 2% over the descent. We build them when an application needs constant, silent, electricity-free torque — the Westminster Great Clock and countless workshop demonstration rigs use exactly this principle.
Compound Weight Motor Interactive Calculator
Vary the falling weight, drum size, and two gear stages to see drum torque, speed multiplication, and ideal output torque.
Equation Used
The falling mass creates a nearly constant force, F = m*g. Drum torque is that force multiplied by the drum radius. Each gear stage multiplies output speed, so ideal output torque is divided by the total speed gain.
- Gear stages are ideal and lossless.
- Stage ratios are entered as speed multiplication ratios.
- Cord does not slip on the drum.
Operating Principle of the Compound Weight Motor
The mechanism is simple in concept and unforgiving in detail. A weight hangs from a cord or chain that wraps a drum. The drum is geared — through a compound train of pinions and wheels — to the output shaft. As the weight falls, the cord unwinds, the drum rotates, and the gear train multiplies that slow rotation into useful speed at the output. The compound part of the name refers to the gear stages, not the weights themselves. You stack two or three reduction stages to convert a 1 mm/s descent into a few hundred RPM at the escapement or whatever load you are driving.
Why build it this way? Because gravity gives you genuinely constant force. A 10 kg weight pulls with 98.1 N every second of its fall — no battery droop, no spring relaxation, no pneumatic regulator drift. That constant-torque drive is why weight-driven clock movements held the precision crown for 400 years before quartz arrived. The compound pulley variant doubles down by adding a movable pulley block, so a single weight can drive a longer runtime at half the descent rate, at the cost of half the available torque.
Tolerances will bite you. Drum runout above 0.3 mm causes cord whip that you will hear as a tick-tock variation in a clock movement. Bearing friction above roughly 5% of the weight's pull torque robs the output of useful power and, worse, makes the descent rate vary with temperature as oil viscosity changes. The most common failure is cord stretch — natural fibre cord creeps under sustained load, the weight slowly lowers itself even when the train is locked, and you wake up to a stopped clock. Switch to braided steel cable or aramid cord for any sustained-load gravity-driven motor build.
Key Components
- Driving Weight: The potential energy source. Sized to deliver enough torque at the drum to overcome train friction plus drive the load. A typical longcase clock uses a 4 to 8 kg lead or brass-cased weight; tower clocks run 100 to 500 kg of cast iron.
- Drum (Barrel): The shaft-mounted cylinder the cord winds onto. Diameter sets the torque arm — a 60 mm drum with a 5 kg weight delivers 1.47 N·m at the input. Concentricity must hold to 0.1 mm or you get cyclic torque variation that the escapement reads as positional error.
- Compound Gear Train: Two or three stages of pinion-and-wheel reduction-in-reverse — the slow drum drives faster output. Ratios of 1:60 to 1:3600 are normal. Each stage loses 2 to 5% to friction, so a three-stage train ends up around 85% efficient on clean pivots.
- Movable Pulley Block (when present): Halves the descent rate and halves the available torque, but doubles the runtime for a given weight drop. The Huygens compound weight movement of 1657 used this trick to reach 7-day operation in a domestic case.
- Cord or Chain: Transfers force from weight to drum. Aramid or steel cable for serious builds — natural gut and silk creep under load and cause descent-rate drift. Chain on a fusee sprocket eliminates wrap-diameter change as the line pays out.
- Maintaining Power Spring: A small pre-loaded spring that drives the train during winding so the clock does not stop or run backwards. Harrison's bolt-and-shutter design from 1722 is the classic implementation.
- Output Escapement or Load: The thing the motor drives. In a clock, an anchor or deadbeat escapement gates the energy. In a Foucault pendulum drive or a museum demonstration hoist, the output is a continuous-load coupling.
Where the Compound Weight Motor Is Used
Weight-driven motors show up wherever you need silent, constant, electricity-independent power for long durations. They are not antiques — modern museum installations, off-grid mechanical sequencers, and educational kinetic sculpture all use them. The reason is the same as it was in 1300: a hanging mass is the most reliable energy storage device humans have ever built.
- Horology: The Westminster Great Clock (Big Ben) drives its hands using three weight motors totalling roughly 1.2 tonnes of cast iron, rewound twice weekly.
- Museum Restoration: The Salisbury Cathedral clock, dated to around 1386 and still operating after restoration, uses paired stone weights to drive a foliot escapement.
- Kinetic Art: Tim Hunkin's Southwold Pier automata use compound weight drives to sequence figure motion over 90-second show cycles without any electrical input to the drive train.
- Education: Physics-classroom Atwood-machine demonstrators built around two weights and a compound pulley illustrate Newton's second law and gear-ratio mechanical advantage in a single rig.
- Off-grid Power: The GravityLight LED lamp, designed by Therefore Product Design in 2013, uses a 12 kg sand-bag weight motor and a compound gear train to drive a small generator for 25 minutes of reading light per lift.
- Industrial Heritage: Restored stamp mills at the Empire Mine State Historic Park in California demonstrate weight-and-cam drives that originally lifted ore-crushing stamps before steam takeover.
- Scientific Instruments: Foucault pendulum installations such as the one at the Panthéon in Paris use a small weight motor coupled through a compound train to maintain swing amplitude against air drag.
The Formula Behind the Compound Weight Motor
The output power and runtime of a compound weight motor depend on three things: the weight, the drop height, and the time over which it descends. At the low end of a typical clock-movement build — a 2 kg weight dropping 1 m over 8 days — you get roughly 28 µW at the drum, which sounds laughable but is exactly what an escapement needs. At the nominal mid-range — say 5 kg over 1.5 m in 7 days — you produce about 121 µW, plenty for a longcase movement with a heavy pendulum. Push to the high end — 200 kg over 4 m in 24 hours, like a tower-clock motor — and you generate roughly 90 W, enough to drive 4 m diameter hands against wind loading. The sweet spot for most workshop and educational builds sits between 1 and 50 W average output.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Pavg | Average mechanical power delivered over the descent | W | ft·lbf/s |
| m | Mass of the driving weight | kg | lb |
| g | Gravitational acceleration (9.81 on Earth) | m/s² | ft/s² |
| h | Total descent height of the weight | m | ft |
| t | Runtime — time taken for the weight to fully descend | s | s |
| η | Train efficiency — multiply Pavg by this for usable output | dimensionless | dimensionless |
Worked Example: Compound Weight Motor in a kinetic sculpture exhibit drive
A gallery installation in Glasgow needs a silent, electricity-free drive for a brass kinetic sculpture that rotates a 600 mm diameter cam disc once every 4 minutes. The curator wants 7-day autonomy between rewinds, a 2.5 m available drop height in the plinth, and a gear-train efficiency of 80%. Pick the weight and check the output power across the typical operating range.
Given
- h = 2.5 m
- t = 604,800 (7 days) s
- η = 0.80 —
- Pload = 0.5 W (estimated cam drag at 1 RPM)
- g = 9.81 m/s²
Solution
Step 1 — work out the input power needed at the drum, accounting for the 80% efficiency:
Step 2 — solve the formula for the nominal weight, given a 7-day runtime over 2.5 m:
That 15 kg is the design centre. A cast-iron cylinder 80 mm diameter by 310 mm long lands you almost exactly there.
Step 3 — check the low end of the typical operating range. If the curator changes their mind and wants 14-day autonomy on the same drop height:
The weight doubles. At 30 kg you start running into plinth-loading and rewinding-effort problems — a curator winding a 30 kg weight by hand with a crank handle is a back-injury claim waiting to file itself.
Step 4 — check the high end. If the runtime drops to 24 hours (daily rewind, common in pre-electric public clocks):
A 2 kg weight is trivial to lift and the gear-train torque is comfortable, but you lose the headline feature — weekly autonomy. The 7-day, 15 kg point is the genuine sweet spot for this gallery brief.
Result
The nominal weight comes out to 15. 4 kg for 7-day silent operation. That mass at 2.5 m drop gives the curator a once-a-week rewind ritual that fits the exhibit's character without straining whoever does the winding. The 14-day version doubles the weight to 30.8 kg and creates real handling problems, while the 1-day version drops to 2.2 kg but kills the autonomy story — the 15 kg, 7-day point is clearly the sweet spot. If you build it and the cam stalls before the weight reaches the floor, suspect: (1) cord stretch on natural-fibre line, where 5 mm of creep per day robs the runtime budget, (2) pivot friction climbing above 25% as winter temperatures thicken the gear-train oil, or (3) drum wrap-diameter change as a thick cord layers over itself, which raises effective torque demand near the top of the wind.
When to Use a Compound Weight Motor and When Not To
Weight motors are not always the right answer. Mainspring drives are smaller and portable, electric motors give you indefinite runtime, and pneumatic drives shrug off dust and heat. Here is how the compound weight motor actually compares on the dimensions that matter for selection.
| Property | Compound Weight Motor | Mainspring (Going Barrel) | Electric Gearmotor |
|---|---|---|---|
| Torque variation across runtime | < 2% (constant gravity) | 30 to 50% (spring uncoils) | < 1% (regulated drive) |
| Typical runtime per wind/charge | 1 to 8 days | 30 to 40 hours | Unlimited (mains) or 4 to 24 h (battery) |
| Power output range | 10 µW to 100 W | 1 µW to 10 W | 1 W to 100 kW+ |
| Vertical space requirement | 1 to 5 m drop needed | Compact, fits a wristwatch | Compact, any orientation |
| Maintenance interval | 10 to 20 years (pivots, cord) | 5 to 7 years (spring fatigue) | 5,000 to 50,000 hours (bearings) |
| Lifespan of energy source | Indefinite (gravity does not wear) | Spring fatigues, replace every 50 years | Motor windings 20,000+ hours |
| Capital cost (workshop build) | £100 to £400 | £40 to £150 | £25 to £200 |
| Ambient noise during operation | Silent except escapement tick | Silent | Audible motor hum and gear noise |
Frequently Asked Questions About Compound Weight Motor
That's almost always cord layering on the drum. As the cord pays out, the wrap diameter at the drum decreases — sometimes by 1 to 2 mm — which reduces the input torque to the train. With less torque the escapement amplitude drops slightly and the rate slows.
Fix it by switching to a fusee-and-chain system or by using a single-layer drum long enough to hold the full descent in one wrap. A 60 mm drum should be at least as long as h × C / (π × D), so all turns sit side-by-side rather than stacking.
Use a compound pulley (movable block) when the available drop height is less than half the runtime your application demands, AND the load can tolerate half the input torque. The trade is exact: 2:1 pulley doubles runtime, halves torque.
Don't use it when the gear train is already at 80%+ efficiency loss — adding pulley friction can push total losses past 35% and your output goes weak. For most longcase clocks the direct drive on a long cord wins.
The first suspect is pivot-hole wear, not friction. Worn pivots let the arbors run eccentric, which causes the gear teeth to mesh at varying centre distances and lose 10 to 15% per stage on a three-stage train. Inspect the pivot holes under 10× magnification — if they look oval rather than round, rebush them.
Second suspect is cord wrap-angle on the drum. If your cord enters the drum at an angle rather than tangentially, it climbs over previous wraps and burns energy as friction. A fairlead pulley fixes this.
Third, check that the maintaining-power spring isn't dragging continuously — it should only engage during winding.
Yes, but the numbers get hostile fast. A 1 kW continuous output for 24 hours requires roughly 86 MJ of stored energy. At a realistic 5 m drop that's 1,750 kg of mass — almost two tonnes hanging in your shaft. The drum bearings have to take that load plus dynamic factors, the rewind mechanism needs a motor of its own, and the safety case for the falling weight becomes the dominant design problem.
Above roughly 100 W average, weight motors stop being practical and you should switch to either a hydraulic accumulator or a battery-electric system with a generator. The historical ceiling of around 1 kW belonged only to public tower clocks where the building itself provided the drop height.
Almost always one of two issues. First — and most embarrassingly common — the gear train is binding because two arbors aren't parallel. Even 0.5° of misalignment over a 200 mm centre distance causes tooth-end loading that doubles friction. Check parallelism with a dial indicator before assuming anything else.
Second is the escapement entry-pallet drop. If the pallet doesn't engage the escape wheel cleanly, the train runs free for a few seconds, the weight accelerates, and then the pallet locks hard and stalls everything. Adjust the pallet depth so engagement happens with the escape wheel tooth landing on the impulse face, not the locking corner.
It needs to fall freely. Any contact between the weight and a guide tube introduces variable friction that destroys the constant-torque advantage you chose this drive for in the first place. A weight brushing the inside of a 50 mm guide tube can lose 5 to 20% of its driving force depending on humidity and dust.
If you must constrain lateral motion — for example to keep the weight from swinging in a tall cabinet — use a single tensioned guide wire running through a small ring on top of the weight, not a tube. Friction stays under 1% of pull and stays consistent across the descent.
Don't use natural-fibre cord for any sustained-load weight motor build. Silk and gut creep under 2 to 5% over the first month of loading, which on a 2.5 m drop costs you 50 to 125 mm of effective travel and several hours of runtime.
Use 7×19 stainless steel cable for weights above 5 kg, or braided aramid (Kevlar) cord for lighter, quieter setups. Both creep less than 0.1% under sustained load. Pre-tension new cable at 110% of operating load for 24 hours before final installation to settle any constructional stretch.
References & Further Reading
- Wikipedia contributors. Weight-driven clock. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
