A conical pendulum is a mass swung on a cord or rod so that the bob travels in a steady horizontal circle rather than swinging back and forth in a plane. Christiaan Huygens analysed it in 1659 and William Bond built it into his 1858 sidereal clock for Harvard College Observatory. The bob's weight and the cord tension combine to supply continuous centripetal force, giving smooth rotary motion at a fixed period. That smoothness is why it drives equatorial telescopes and centrifugal governors silently — no tick, no impulse, no escapement jolt.
Conical Pendulum Interactive Calculator
Vary target rotation period, cone angle, and gravity to size the vertical height, cord length, orbit radius, and rpm while the pendulum diagram animates.
Equation Used
The article period relation says a conical pendulum depends on vertical height h, not directly on cord length L. This calculator inverts that relation to find the height needed for a target period, then uses the selected cone angle to calculate cord length and orbit radius.
- Steady circular motion with the cord at a fixed cone angle.
- Cord or rod is massless, rigid, and measured from the pivot to the bob center.
- Cone angle theta is measured from vertical.
- Friction, air drag, and pivot losses are neglected.
Operating Principle of the Conical Pendulum
Hang a bob from a fixed pivot on a cord of length L, then push it sideways until it sweeps a horizontal circle at a constant angle θ from vertical. Two forces act on it — gravity pulling straight down, and tension pulling along the cord toward the pivot. The horizontal component of tension supplies the centripetal force the bob needs to keep curving. The vertical component balances gravity. Solve those two equations together and the period drops out as T = 2π × √(h / g), where h is the vertical distance from pivot to the plane of the bob — not the cord length. That distinction matters. If you size your drive against L instead of h, your rate is wrong by a factor of √(cos θ).
The mechanism is isochronous in a useful but limited sense. For small cone angles the period barely changes with amplitude, which is why William Bond used it for sidereal timekeeping. Push the angle past about 30° and the geometry stiffens — period shortens noticeably as θ grows because h shrinks. Most precision conical-pendulum clocks run at θ between 3° and 8° to stay in the flat part of the curve.
What fails in practice is the suspension. A flexing bob arm, an off-axis pivot, or a pivot with stiction will turn the smooth circle into a wandering ellipse. You see this as a rate that drifts with bob amplitude. A worn pivot bearing also bleeds energy each revolution, so the drive train has to keep nudging the bob faster — and the faster you nudge, the further off-isochronous you run. Build the pivot stiff, build the arm rigid, and feed energy in tangentially through a low-friction coupling.
Key Components
- Bob: The rotating mass at the end of the cord. Mass is typically 100 g to several kg depending on application — Bond's observatory clock used roughly 1.4 kg. Mass cancels out of the period equation but sets the energy stored per revolution and therefore how tolerant the system is to friction losses.
- Suspension cord or rod: Connects bob to pivot. Length L sets the cone geometry together with angle θ. For precision work the cord must be inextensible — silk, music wire, or a fine steel rod. Stretch under centrifugal load shifts the period and shows up as a rate error of roughly 0.5 s/day per 10 µm of stretch on a 1-metre suspension.
- Pivot or universal joint: Allows the cord to sweep a cone freely without binding. Stiction here is the single biggest cause of period drift. A jewelled pivot or a fine ball-and-socket with surface finish below Ra 0.2 µm gives the cleanest rotation. Friction torque must stay under about 1% of the centripetal moment for a stable rate.
- Drive coupling: Replaces the energy lost per revolution to air drag and pivot friction. In Bond-style clocks this is a tangential lever that nudges the bob arm once per revolution. Modern telescope drives use an electric motor coupled through a worm and wheel. The drive must add energy without disturbing the cone angle, otherwise the bob hunts.
- Frame and base: Holds the pivot rigidly. Any flex in the support translates directly into period error because the effective height h changes. Cast iron or thick steel weldments are typical. Resonance in the support near the rotation frequency (usually 0.5 to 2 Hz) will make the bob swing in figure-8 patterns instead of circles.
Where the Conical Pendulum Is Used
The conical pendulum is the natural choice whenever you need continuous, silent, near-isochronous rotary motion without an escapement's tick. That eliminates impulse jolts that would blur a long-exposure photographic plate or a delicate measurement. It also turns up wherever rotating mass needs to self-regulate against speed changes — the centrifugal governor on a steam engine is mathematically the same device, just driven instead of free-running.
- Astronomical instruments: William Bond & Son's spring-driven sidereal clock at Harvard College Observatory (1858) used a conical pendulum to drive an equatorial telescope smoothly across the sky during long exposures.
- Steam and gas engine control: James Watt's centrifugal governor (1788) is a driven conical pendulum — flyballs swing outward as engine speed rises, lifting a sleeve that throttles the steam valve.
- Precision horology: Reefer Devereux and later Eureka-style clocks built conical-pendulum movements where the pendulum itself drives the gear train, replacing the escapement entirely for silent operation in libraries and bedrooms.
- Physics demonstration: University teaching labs use 1-metre conical pendulums to demonstrate centripetal force, the relationship between cone angle and period, and the limits of isochronism — a standard experiment at MIT and Cambridge undergraduate level.
- Amusement engineering: Swing-ride attractions like the chair-o-plane are scaled-up conical pendulums — the same equation predicts the cone angle riders settle at given the rotation speed of the central tower.
- Industrial sensing: Early flow meters and tachometers used small conical-pendulum elements to convert rotational speed into a measurable cone angle, calibrated against the same T = 2π × √(h / g) relationship.
The Formula Behind the Conical Pendulum
The period equation tells you how long one full revolution takes given the geometry of the cone. What matters in practice is the vertical drop h from pivot to bob plane — not the cord length L. At small cone angles (under 10°) h is almost equal to L, the period barely changes with amplitude, and you have a near-isochronous timekeeper. Open the angle to 30° and h has dropped to 0.87 × L — period shortens by 7%, which on a 1-second pendulum is 6,000 seconds per day. Past 45° the geometry runs away from you fast and the device behaves more like a governor than a clock. The sweet spot for horological work is 3° to 8°.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| T | Period of one full revolution of the bob | seconds (s) | seconds (s) |
| L | Length of the suspension cord from pivot to bob centre | metres (m) | feet (ft) |
| θ | Half-angle of the cone, measured from vertical | degrees (°) | degrees (°) |
| h | Vertical distance from pivot to the plane of the bob (L × cos θ) | metres (m) | feet (ft) |
| g | Gravitational acceleration | 9.81 m/s² | 32.2 ft/s² |
Worked Example: Conical Pendulum in a kinetic art installation rotating drive
A kinetic sculpture studio in Rotterdam is building a slow-rotating overhead mobile for a hotel atrium. The artist wants a 4-second revolution period with no visible drive — a conical pendulum hidden in the ceiling cavity will spin the cable that carries the mobile. Cord length available in the cavity is 1.20 m. You need to set the cone angle that gives exactly T = 4.0 s and check what happens if the pivot rises 10 mm during installation or the artist later asks for a faster 3-second rotation.
Given
- T = 4.0 s
- L = 1.20 m
- g = 9.81 m/s²
Solution
Step 1 — rearrange the period equation to solve for the required vertical drop h at the nominal 4-second period:
That h is way larger than the 1.20 m cord — meaning a 4-second period is geometrically impossible with a 1.20 m suspension. The longest period a 1.20 m cord can deliver is when θ → 0 and h → L, giving:
So the artist's 4-second target is off the table without lengthening the cord. The practical operating range for this 1.20 m cord runs from about T = 2.0 s (small cone angle, near-vertical) down to T = 1.6 s at θ = 45° where h = 0.849 m:
Step 2 — pick a workable nominal. A 2.0-second revolution at θ ≈ 6° is the sweet spot — slow enough that the mobile reads as drifting rather than spinning, and isochronous enough that small drive perturbations don't visibly change the rate:
Wait — that's not 6°, that's 34°. The 1.20 m cord forces a wide cone to hit even 2.0 s. At the low end of the usable range (θ = 5°, h = 1.196 m) the period is T = 2.194 s — only 9 ms slower than θ → 0. At the high end (θ = 45°, h = 0.849 m) the period drops to 1.85 s and the bob is tracing a circle of radius 0.849 m, which won't fit in most ceiling cavities.
Step 3 — quote the new design back to the artist. With a 1.20 m cord, achievable periods run 1.85 s to 2.20 s. Hitting 4 s requires extending the cord to roughly 4.0 m (h ≈ 3.97 m at small angle), which means dropping the pivot through the ceiling slab.
Result
The 4-second target is not achievable with a 1. 20 m cord — the geometric ceiling is 2.20 s, so you go back to the artist with a choice between a 2-second period or a 4-metre suspension. At the low end of the useful range (θ = 5°) the period sits at 2.19 s and the bob barely tilts, giving a tight 10 cm circle ideal for a small ceiling cavity. At the nominal θ = 34° the period is 2.0 s but the bob now traces a 0.68 m radius circle, so the cavity has to be at least 1.5 m wide. If your measured period drifts from the predicted value, suspect three things in order: (1) pivot stiction from a dirty or undersized bearing — the cone wanders into an ellipse and the average period grows; (2) cord stretch under centripetal load if you used cotton or nylon instead of music wire — 0.5 mm of stretch on this 1.20 m cord shifts the period by ~5 ms; (3) frame flex at the pivot mount, which lowers the effective h and shortens the period by 1-2% per millimetre of vertical sag.
When to Use a Conical Pendulum and When Not To
The conical pendulum competes against the standard plane pendulum and the balance wheel for any application that needs a periodic mechanical reference. Each wins on different axes — pick based on what your application actually demands.
| Property | Conical Pendulum | Plane Pendulum (anchor escapement) | Balance Wheel & Hairspring |
|---|---|---|---|
| Motion type | Continuous rotation, no impulse | Reciprocating swing with tick | Reciprocating oscillation with tick |
| Typical period range | 1–4 s per revolution | 0.5–2 s per swing | 0.2–0.4 s per oscillation |
| Isochronism (rate vs amplitude) | Excellent below θ = 10°, degrades fast above 30° | Excellent below 5° swing (circular error) | Excellent across full amplitude with proper hairspring |
| Audible noise | Silent — no escapement | Audible tick at every beat | Audible tick, higher pitch |
| Achievable rate accuracy | 1–10 s/day with care | 0.1–1 s/day in a regulator | 1–10 s/day in a wristwatch, sub-second in a chronometer |
| Suitability for portable use | Poor — requires fixed vertical reference | Poor — gravity-dependent | Excellent — works in any orientation |
| Drive complexity | Simple tangential nudge or motor | Requires escapement | Requires escapement |
| Best application fit | Telescope drives, governors, silent clocks, kinetic art | Tower clocks, regulators, longcase clocks | Watches, chronometers, portable timekeepers |
Frequently Asked Questions About Conical Pendulum
The bob is picking up a small in-plane oscillation on top of the rotation, and the two motions combine into an ellipse. The usual cause is asymmetry in the drive — a tangential lever that nudges harder on one side, or a motor coupling that has 0.1 mm of radial runout. Each revolution adds a tiny lateral kick at the same point in the cycle, which builds up coherently into the elliptical mode.
Diagnose by killing the drive and watching the freewheel decay. If the cone stays circular as it spins down, the drive is the culprit. If the ellipse persists during decay, your pivot has directional stiffness — typically a knife-edge or jewel that's seated unevenly. Re-bed the pivot and the ellipse usually disappears.
2.5% is exactly the kind of error you get when you measured cord length L and used it directly in the equation instead of computing h = L × cos θ. At θ = 13° you'd be off by exactly cos(13°) = 0.974, which lengthens the predicted period by 1.3% — close to half your error. The other half is usually pivot height: people measure L from the top of the pivot housing, but the effective pivot point is the centre of the ball or the knife-edge, often 5–10 mm lower.
Re-measure h directly with a ruler from the actual pivot contact point down to the bob centre during rotation. The number almost always reconciles within 0.5%.
Conical, almost always — that's the whole reason the type exists. A plane pendulum needs an escapement, and every escapement makes noise. Even the quietest deadbeat tick is audible at 3 metres in a quiet room. A conical pendulum with a worm-and-wheel drive runs silently because there's no impulse event — the drive feeds energy continuously.
The trade is rate stability. A good conical pendulum holds 5–10 s/day; a regulator-grade plane pendulum holds 0.1–1 s/day. If the clock just needs to look right and stay within a minute over a week, conical wins. If it has to keep observatory time, build the plane pendulum and put it in another room.
Around 30°. Below that, dT/dθ stays small and the period is reasonably amplitude-independent — perturb the bob by 1° and the rate barely shifts. Above 30°, dT/dθ grows fast because h = L × cos θ falls steeply. At 45° a 1° amplitude wobble changes the period by about 0.4%, which is 350 s/day — useless for timekeeping but exactly what you want in a Watt governor, where you need the cone angle to respond strongly to speed changes.
If your application needs both — a stable rate and a wide angle — you're using the wrong mechanism. Switch to a plane pendulum with deadbeat escapement.
Extra drive power widens the cone angle, which lowers h, which shortens the period — so the clock should run fast, not slow. If you're seeing it run slow with more drive, the drive coupling is dragging on the bob arm during part of each revolution. The friction from the drag bleeds energy back out, but it also applies a tangential retarding force at a fixed phase, which behaves like a soft brake.
Check the drive contact geometry — a tangential lever should kiss the bob arm and release cleanly. If it stays in contact for 20° of arc instead of a brief impulse, you'll see this exact symptom. Shorten the engagement or switch to a non-contact magnetic drive.
No, and this is the fundamental reason marine chronometers used balance wheels instead. A conical pendulum needs a fixed vertical gravity reference. Any platform tilt shifts the apparent vertical, and the cone goes elliptical or precesses. Linear acceleration adds vector to gravity and changes the effective g, which directly changes the period through T = 2π × √(h / g).
For portable or moving applications you want a balance wheel and hairspring, which work in any orientation because the restoring force comes from the spring rather than gravity.
References & Further Reading
- Wikipedia contributors. Conical pendulum. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
