How a Mechanical Watch Works

A mechanical watch is, at its simplest, a machine that does three things: it stores mechanical energy, it releases that energy in small, evenly counted portions, and it uses the counted releases to move hands around a dial. Every mechanical watch or clock, regardless of age, complexity or country of origin, can be understood as a solution to those three problems, which makes the overall architecture far more approachable than the number of tiny parts might suggest.

Energy storage is usually a coiled mainspring housed in a barrel, though weight-driven clocks store energy by raising a mass against gravity instead. The mainspring is wound by hand (via the crown and winding stem in a modern watch, or a key in older examples) or, in self-winding watches, by an oscillating weight driven by the wearer's motion. As the spring unwinds, it turns the barrel, which meshes with the first wheel of the gear train.

The gear train — typically centre wheel, third wheel and fourth wheel in a watch — is a set of stepped-up gears that both transmits the barrel's torque and increases rotational speed at each stage, ending at the escape wheel. Left alone, a wound spring would simply run down as fast as friction allowed, releasing all its energy in a few seconds. The escapement exists to prevent that: it locks the escape wheel, releases it briefly to let the train advance by a fixed, small amount, and locks it again, over and over, in a rhythm set by an oscillator.

The oscillator is a balance wheel and hairspring in a watch, or a pendulum in most clocks. It swings back and forth at a highly consistent rate determined by its moment of inertia and the stiffness of its spring (or, for a pendulum, by its effective length and gravity). Each swing releases one tooth of the escape wheel through the escapement, and in return the escapement gives the oscillator a small push, called impulse, to keep it swinging despite friction and air resistance. This exchange — the oscillator controlling the escapement's release, and the escapement replenishing the oscillator's energy — is the heart of every mechanical timekeeper.

Finally, the counted releases are converted into a visible time display through the motion work: a further small gear train, geared down from the fourth wheel through the cannon pinion, that turns the minute and hour hands at the correct relative speeds. A watch that is running is therefore always doing the same underlying thing regardless of its outward complexity: converting stored energy into a train of tooth releases at a rate governed by an oscillator, and translating that count into hand positions.