An Arnold-type spring-detent escapement lets the balance run freely between brief interactions with the escape mechanism. A locking detent releases the escape wheel on one swing, the wheel delivers an impulse to the balance, and the return swing passes without a second release. The animation follows that sequence, while the calculator counts tooth releases and compares entered angular spans.
Arnold Chronometer Escapement Interactive Calculator
Vary balance frequency, escape-wheel tooth count, and impulse arc to see daily tooth release and detached running geometry.
Equation Used
- One escape tooth is released per full balance oscillation.
- Frequency is in full cycles per second, not beats per second.
- The total entered arc is peak-to-peak balance angle. Contact share is an angular fraction.
The drawing shows one-way release and a flexible return pass. Tooth and jewel profiles are illustrative, and the calculation does not predict clock rate error.
Unlock, impulse and a free return
The balance and its spring form the oscillator. A jewel on the detent locks the powered escape wheel between releases. During the impulse-direction swing, the balance’s discharge feature acts through the passing spring and moves the detent clear of the locking tooth.
The escape wheel then advances and delivers energy through the impulse pallet or jewel on the balance roller. The detent returns to lock the wheel. On the return swing, the flexible passing spring yields so the balance can pass without unlocking the wheel again. This gives one release per full oscillation in the model.
Parts in the drawing
- Balance and hairspring: oscillate about the balance staff.
- Impulse roller and jewel: receive the escape-wheel impulse.
- Discharge feature: initiates unlocking on the appropriate swing.
- Spring detent and locking jewel: hold and release the escape wheel.
- Passing spring: flexes during the return pass.
- Escape wheel: advances one tooth for each complete balance oscillation.
The view explains the sequence. It does not specify working tooth profiles, jewel positions, spring dimensions or an allowable balance amplitude. The spring detent’s elastic support and the flexible passing spring perform different functions.
Marine and pocket chronometers
Royal Museums Greenwich documents surviving marine and pocket chronometers with Arnold-type spring-detent escapements. Its record for marine chronometer ZAA0007 identifies the escapement and fusee movement; pocket chronometer 2168 provides another example of the spring-detent arrangement.
Escapement type alone does not guarantee a daily rate. Balance and spring behavior, power delivery, position, temperature, adjustment and condition all contribute to timekeeping. A geometric animation cannot certify those properties.
The historical Arnold and Earnshaw designs also have distinctions that should not be inferred from a maker name alone. A particular instrument’s collection record or technical drawing is the appropriate reference for identifying its mechanism.
Tooth counting and angular comparisons
Let f be full balance oscillations per second and N be escape-wheel teeth. With one tooth released per full oscillation, Teeth per Day = 86400f and Wheel Rev/Day = 86400f/N. Mean wheel speed in rpm is 60f/N, although the wheel moves intermittently.
Two beats make one full oscillation. Thus beats per hour = 7200f. A rate of 14400 beats per hour means 2 full oscillations per second, not 4. At 4 full oscillations per second, the beat rate is 28800 per hour.
The angular tiles use peak-to-peak span θ and entered impulse arc θi. Detached Arc = max(0, θ − θi). Contact Share = 100θi/θ, limited to 0–100%. These compare the impulse arc with one peak-to-peak span; they do not measure actual total contact throughout a cycle.
A full oscillation travels out and back, covering twice the peak-to-peak span. Balance speed also varies during a swing. Therefore Contact Share is neither a fraction of total angular distance per full cycle nor a percentage of elapsed time in contact. Unlocking and the flexible return contact are not included in the tile arithmetic.
The frequency and arc inputs are independent. Changing an arc does not alter daily release count or simulate isochronism, rate error, shock response or tripping.
Worked example: distinguish cycles from beats
At N = 15 and f = 4 full oscillations/s, Teeth per Day = 4 × 86400 = 345600. Wheel Rev/Day = 345600/15 = 23040, equivalent to a mean 16 rpm. The associated beat rate is 28800 beats per hour.
With peak-to-peak span θ = 360° and impulse arc θi = 40°, Detached Arc is 320.0° and Contact Share is 11.1111%, displayed as 11.1%. This 360° peak-to-peak span corresponds to 180° amplitude from the central position. The two-direction travel during a full oscillation is 720°.
At f = 2, daily tooth count halves to 172800 and wheel revolutions to 11520, or mean 8 rpm. This corresponds to 14400 beats per hour. The angular tiles remain unchanged.
At N = 30 and f = 4, daily tooth count stays 345600 while wheel revolutions become 11520. Changing θi to 60° at θ = 360° gives Detached Arc 300.0° and Contact Share 16.7%, without changing either counting result. These comparisons do not predict a real chronometer’s daily accuracy or determine its complete gear train.
Comparing escapement actions
| Arrangement | Characteristic action |
|---|---|
| Spring detent | Unlocking and impulse occur on one swing; the return passes without a second wheel release. |
| Lever escapement | A lever alternates locking and impulse interactions on successive swings. |
| Verge escapement | Alternating pallets on the verge engage the crown wheel. |
This motion comparison assigns no universal daily accuracy, service interval or shock rating. Contact geometry and the entire oscillator system matter to performance.
Detent counting questions
Full oscillations per second. Each comprises two swings or beats. Multiply the entered value by 7200 for beats per hour.
At fixed f, the same number of teeth is released each day. More teeth per revolution means fewer revolutions for that count.
No. It is impulse arc divided by one peak-to-peak span. Speed varies during the swing, and the calculator does not integrate contact time.
No. It subtracts impulse arc from the entered span. Unlocking and return-spring interaction are not included in that arithmetic.
No. Frequency is an independent input. No spring, temperature, energy-loss or isochronism model changes it.
Not necessarily. Peak-to-peak span runs from one extreme to the other. An amplitude measured from the central position is half that span; measurement conventions must be checked.
No. Profiles and contacts are schematic. Dimensions, material and adjustment procedures require the specific escapement’s technical documentation and condition assessment.
No. Actual rate is measured on the complete instrument under specified conditions. The event counter does not certify accuracy or diagnose a timing fault.
Historical examples
Royal Museums Greenwich: marine chronometer ZAA0007 and pocket chronometer 2168. Both records identify Arnold-type spring-detent arrangements.
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.