Bloxam's Gravity Escapement is a precision pendulum escapement, also called the Three-pin pallet gravity escapement, in which two pivoted gravity arms — not the going train — deliver impulse to the pendulum. Andrew Bloxam, an English clergyman-horologist, described it in 1853, predating Edmund Beckett Denison's better-known double three-legged variant. The escape wheel only lifts the arms; the arms then fall under their own weight to push the pendulum. This isolation from train torque produces sub-second-per-week accuracy in tower clocks where wind load on the hands swamps lighter escapements.
Bloxam's Gravity Escapement Interactive Calculator
Vary gravity-arm mass, center-of-mass radius, lift angle, and beat rate to see the delivered pendulum impulse energy and animated escapement motion.
Equation Used
The calculator treats each gravity arm as a mass whose center of mass is lifted through a small angle. The stored gravitational energy, m*g*r*(1 - cos(theta)), is the ideal impulse energy available to the pendulum on that beat.
- One raised gravity arm delivers one impulse per beat.
- Arm center of mass moves on a circular arc about its pivot.
- Friction, rebound, and impact losses are neglected.
- Lift angle is small and measured from the dropped arm position.
How the Bloxam's Gravity Escapement Actually Works
The Bloxam's Gravity Escapement, also called the Three-pin pallet gravity escapement, works by separating two jobs that most escapements lump together: locking the train and impulsing the pendulum. In a deadbeat or anchor escapement, every variation in train torque — a sticky pinion, a gust of wind on the minute hand, oil thickening in winter — feeds straight into the pendulum and shifts its rate. Bloxam's solution puts a small lightweight gravity arm on each side of the pendulum. The escape wheel, a 3-pin lifting wheel running through a count train, only has to raise each arm by a fixed angle, typically 2.5° to 3°. Once raised, the arm drops onto the pendulum during the next swing and impulses it with a force determined by gravity alone — not by the wheel train.
The geometry is unforgiving. The lifting pins must be spaced exactly 120° apart on the wheel, and the locking faces on the arms must match within roughly 0.02 mm or one arm receives more lift than the other and the pendulum picks up a positional error you can hear as an uneven tick. If the gravity arms are too heavy you overdrive the pendulum, increase amplitude, and pull the rate via circular error. Too light and the arm fails to clear the locking detent on a cold morning when pivot oil has thickened. Common failure modes are tripping (the wheel runs free because lock depth fell below 0.3 mm of engagement), bouncing arms (worn pivot holes letting the arm rebound and double-impulse), and butting (escape wheel pin striking the arm face instead of sliding under it — caused by an arm that drops too slowly).
The practical outcome is that load fluctuations on the going train, which would shift a deadbeat clock by several seconds a day under wind load, shift a Bloxam clock by milliseconds. The pendulum sees a constant impulse, every swing, every season.
Key Components
- Three-pin escape wheel: A lightweight wheel carrying 3 lifting pins at 120° spacing. Its only job is to raise the gravity arms; it never touches the pendulum directly. Pin position tolerance is around ±0.05 mm — beyond that, alternating ticks become unequal and rate stability suffers.
- Left and right gravity arms: Two pivoted arms with calibrated weights, typically 8 to 25 grams in tower-clock scale. Each arm rises about 2.5° during lift and falls under gravity to deliver impulse to the pendulum. Mass match between arms must be within 0.5% to keep the tick even.
- Locking detents (pallets): Hardened steel faces at the top of each arm that catch a pin on the escape wheel and stop the train. Engagement depth runs 0.3 to 0.5 mm. Less than 0.3 mm and the wheel trips through under shock load; more than 0.5 mm wastes lift angle.
- Pendulum impulse pin: A polished pin or roller on the pendulum rod that the gravity arm bears on during impulse. Surface finish below Ra 0.2 µm matters — a rough pin causes arm bounce and chirpy ticks.
- Fly fan or count train: A small air-braked fan downstream of the escape wheel limits the wheel's rotational speed during the brief unlocked phase, so pins don't crash into the next arm. Without it the escapement self-destroys within hours.
Who Uses the Bloxam's Gravity Escapement
Bloxam's escapement and its descendants live almost exclusively in high-stakes pendulum clocks — public, scientific, and astronomical — where the cost of an extra escape wheel and pair of gravity arms is trivial against the cost of being wrong. The Three-pin pallet gravity escapement appears wherever wind, temperature, or train friction would otherwise corrupt rate.
- Public horology: Tower clocks following the lineage of the Westminster Great Clock at the Palace of Westminster — Big Ben uses the closely related Denison double three-legged gravity escapement, directly inspired by Bloxam's 1853 design.
- Astronomical observatory clocks: Mid-19th century transit-room regulators at observatories such as Greenwich and Edinburgh used gravity escapements where pendulum amplitude had to stay constant to within fractions of a degree to keep circular-error rate shifts below 0.1 second per day.
- Civic clock restoration: Smith of Derby and Thwaites & Reed routinely service Bloxam-pattern movements in British town hall clocks built between 1855 and 1910.
- Scientific instrument preservation: Museum-grade regulators at the Royal Observatory and the Science Museum in London are kept running on their original gravity escapements rather than retrofitted with quartz drives, because the mechanism itself is the artefact.
- Modern precision pendulum clockmaking: Independent makers such as Buchanan Clocks in Australia and the late David Walter built bespoke gravity-escapement regulators specifying Bloxam-type three-pin geometry for collectors.
- Education and demonstration: British Horological Institute teaching benches use full-scale Bloxam mock-ups to show students how impulse isolation works before moving on to chronometer-grade detents.
The Formula Behind the Bloxam's Gravity Escapement
What you actually need to compute is the impulse energy delivered to the pendulum per beat — because that energy, divided by the pendulum's stored energy, sets the Q-factor and therefore the timekeeping. Below the typical sweet spot of 2.5° to 3° lift angle and 8 to 25 g arm mass, the arm fails to overcome pivot friction on cold mornings and the clock stops. Above that range you overdrive amplitude, push circular error past useful limits, and lose rate stability. The formula tells you where in that band a given build sits.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Eimp | Impulse energy delivered per beat | J (joule) | ft·lbf |
| marm | Effective mass of the gravity arm at its centre of gravity | kg | lb |
| g | Gravitational acceleration | 9.81 m/s² | 32.2 ft/s² |
| Larm | Distance from pivot to arm centre of gravity | m | in |
| θlift | Angle through which the arm is raised by the escape wheel pin | rad (or deg) | deg |
Worked Example: Bloxam's Gravity Escapement in a municipal town hall tower clock
A municipal restoration project in a 1878 town hall tower clock requires verification of the Bloxam-pattern escapement before reinstalling the 1.5 m seconds pendulum. The arms measure 18 g effective mass, the centre of gravity sits 95 mm from the pivot, and the design lift angle is 2.8°. You need to confirm impulse energy lands inside the workable band so the pendulum holds 1.5° amplitude under typical wind loading on the 1.2 m hands.
Given
- marm = 0.018 kg
- Larm = 0.095 m
- θlift,nominal = 2.8 deg
- g = 9.81 m/s²
Solution
Step 1 — convert nominal lift angle to radians and compute the cosine term:
1 − cos(2.8°) = 1 − 0.99881 = 0.00119
Step 2 — compute nominal impulse energy at 2.8° lift:
Step 3 — check the low end of the typical operating range, 2.0° lift, which is what you might see if locking depth has crept up through pallet face wear:
Elow = 0.018 × 9.81 × 0.095 × 0.000609 = 1.022 × 10−5 J ≈ 10 µJ
That is roughly half the nominal energy. The pendulum will still run but amplitude collapses to around 1.0°, and on a cold January morning when the pivot oil thickens you'll see the clock stop entirely — a classic symptom of a tower clock that runs all summer and dies every winter.
Step 4 — check the high end, 3.5° lift, which happens when an over-eager restorer files the locking face too aggressively:
Ehigh = 0.018 × 9.81 × 0.095 × 0.00187 = 3.137 × 10−5 J ≈ 31 µJ
Amplitude climbs past 2°, circular error grows, and rate becomes sensitive to barometric pressure changes you previously didn't see.
Result
Nominal impulse energy is approximately 20 µJ per beat, which puts the pendulum at the design 1. 5° amplitude with a comfortable margin against wind-loaded hand torque. At 2.0° lift the energy halves to 10 µJ and you risk winter stoppage; at 3.5° lift it climbs to 31 µJ and the clock becomes a barometer. The sweet spot is narrow — a few tenths of a degree — which is why Bloxam-pattern escapements demand careful pallet adjustment. If your measured amplitude is below predicted, suspect: (1) hairline corrosion on the escape wheel pins increasing lift friction, (2) gravity arm pivots running dry — clock oil viscosity above 200 cSt at winter temperature can swallow 30% of the impulse, or (3) a bent fly fan arbor letting the wheel overrun and rob lift travel.
Bloxam's Gravity Escapement vs Alternatives
Bloxam's gravity escapement competes with three other approaches in precision pendulum clocks. Each makes a different trade between train-torque immunity, complexity, and cost.
| Property | Bloxam's Gravity Escapement (Three-pin pallet) | Graham Deadbeat Escapement | Denison Double Three-Legged Gravity |
|---|---|---|---|
| Rate accuracy under variable train torque | ±0.1 sec/day typical | ±2 to 5 sec/day under wind load | ±0.05 sec/day typical |
| Complexity (part count) | Moderate — escape wheel, 2 arms, fly | Low — escape wheel, anchor only | High — 2 escape wheels, 2 arms, fly |
| Cost to build (skilled hours) | ~80 to 120 hours bench time | ~30 to 50 hours bench time | ~120 to 180 hours bench time |
| Pendulum amplitude required | 1.0° to 2.0° | 2° to 4° | 1.0° to 1.5° |
| Sensitivity to barometric pressure | Low if amplitude held | Moderate | Very low |
| Best application fit | Tower and civic clocks | Domestic and shop regulators | Top-tier public clocks (Big Ben class) |
| Service interval | 5 to 10 years | 3 to 5 years | 5 to 10 years |
Frequently Asked Questions About Bloxam's Gravity Escapement
Equal arm mass is necessary but not sufficient. The two failure points are usually pivot friction asymmetry — one arm's pivot hole has worn oval while the other has not — and unequal lift-pin contact geometry on the escape wheel. Check arm fall time with a slow-motion camera: each arm should reach the impulse pin in the same number of milliseconds. If one arm is consistently slower by even 5 ms, you'll hear it as a limping tick.
The fix is rarely to add weight. Re-bush the slow pivot first, then verify the lift pin on that side isn't sitting 0.05 mm proud of the others.
Choose Bloxam when you want gravity-escapement immunity to train torque without the fabrication cost of two escape wheels meshed at 90°. For a town hall clock or a private regulator where ±0.5 second per week is good enough, Bloxam delivers it for roughly two-thirds the bench hours. Reserve Denison's design for projects where you need ±0.05 sec/day and the budget supports it — Big Ben, observatory regulators, top-tier independent clockmaker showpieces.
Bloxam also wins on serviceability. With a single escape wheel the count train alignment is one job, not two, and a future restorer 80 years from now will thank you.
Work backwards from the energy lost per beat. A typical seconds pendulum bob of 6 to 10 kg suspended on a steel rod loses around 15 to 30 µJ per beat to air drag and suspension flex. Match your impulse energy to that loss plus 20% margin, then back-solve the formula for arm mass given a chosen lift angle of 2.5° to 3° and a pivot-to-CG distance set by your geometry.
Rule of thumb: for a 1 m seconds pendulum with a 7 kg bob, an 18 g arm at 95 mm Larm and 2.8° lift is a safe starting point. Build it adjustable — slot the arm so you can shift the CG by ±5 mm — because no calculated value survives first contact with real pivot friction.
Almost certainly. Standard clock oil viscosity roughly doubles between 20°C and −5°C. That extra drag at the gravity-arm pivots eats directly into the already-small impulse energy, which is only around 20 µJ in a typical Bloxam build. Once viscous drag exceeds about 30% of nominal impulse, amplitude collapses and the pendulum drops below the lock-clearing threshold.
The diagnostic is to put a thermometer next to the movement and log stop events against temperature. If failures correlate below a threshold temperature, switch to a synthetic clock oil rated for low-temperature service, or insulate the case. Adding arm mass is a last resort because it shifts circular error.
Yes. The three-pin pallet gravity escapement is the descriptive name for the geometry Andrew Bloxam published in 1853 — three lifting pins on the escape wheel acting against two pivoted gravity arms (pallets). Some 19th-century texts call it Bloxam's, others call it the three-pin pallet design; the mechanism is identical.
Don't confuse it with Denison's double three-legged gravity escapement, which uses two separate three-leg wheels rather than a single three-pin wheel. Denison developed his design partly in response to perceived weaknesses in Bloxam's, but both share the gravity-impulse principle.
The formula gives ideal impulse energy assuming the arm transfers all its falling-energy to the pendulum. In practice, three loss paths eat into that: arm-to-pendulum impulse pin friction (worse with rough surfaces above Ra 0.2 µm), arm bounce on a stiff suspension, and energy stored in the impulse pin's elastic deformation if you used a soft material like brass instead of hardened steel.
Measure pendulum amplitude with a protractor scale and back-calculate actual delivered energy. If it's 70 to 80% of theoretical you're in normal territory. Below 60% and you have a mechanical problem — usually a bouncing arm or a soft impulse pin.
You can, but the design becomes fussy. A half-second pendulum needs roughly four impulses per second instead of two, doubling the demand on the count train and the fly fan. Lift angle has to drop to around 2° to keep the arm fall time inside the half-period, and arm mass shrinks proportionally to roughly 8 to 12 g.
Practical builders almost always pair Bloxam's design with a 1-second or 1.5-second pendulum because the arm-fall dynamics fit those periods cleanly. For a half-second pendulum, a deadbeat with a remontoire is usually the better engineering choice.
References & Further Reading
- Wikipedia contributors. Gravity escapement. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
