The pallet fork holds the escape wheel, releases it, and passes an impulse through the balance's roller. Locking, drop and impulse are three separate angles, and all three live in one flick of the fork.
Pallet fork: lock, drop and impulse
issue ISSUE 01checked 2026-09-16responsible Escapement Papers (a named bench, not a person)
tape card: the three angles
| Angle | Typical size | If it is wrong |
|---|---|---|
| Locking | 1-2 degrees | Shallow lock skips; deep lock costs amplitude |
| Drop | 1.5-2.5 degrees | Too much drop wastes impulse |
| Impulse | 2-3 degrees | Too little gives low amplitude |
| Total | 5-7 degrees | The sum is what the pallet geometry has to fit |
Three jobs in one flick
The fork's entry and exit pallets sit in the teeth of the escape wheel. As the balance's roller jewel passes, the fork is nudged, one pallet slides out of a tooth and the other slides in. In that slide the wheel pushes the fork, and the fork pushes the balance. Lock, unlock and impulse happen in a fraction of a beat.
| Action | What the pallet does | What the balance feels |
|---|---|---|
| Locking | Rests against a tooth face | Nothing; the train is held |
| Unlocking | Slides off the locking face | A small loss, subtracted from the impulse |
| Impulse | Takes the tooth's push and passes it on | The energy that keeps the balance swinging |
Why drop is the angle nobody quotes
Drop is the free movement of the escape wheel between one pallet releasing and the other catching. Some drop is necessary, because without it the wheel would be locked at both ends of the motion. Too much drop is a loss, because the wheel moves without doing any work. It is the angle that separates a well-adjusted escapement from a merely working one.
Locking, drop and impulse are conventionally quoted in degrees of escape-wheel rotation, which is why they are comparable across movements with different wheel sizes.
Boundary: the fork is not the roller and not the stud
The roller is the disc on the balance staff that carries the impulse jewel; the fork touches it but does not own it. The stud holds the hairspring's outer end. Both are adjacent to the fork and neither is part of it, and a fault in either produces symptoms that look like a fork fault.
The misreading worth naming: blaming a poor amplitude on the pallet fork when the roller jewel is loose or the stud is rotated. The fork is visible and the roller is small, so the fork gets the suspicion.
A position the bench will defend
The bench holds that total lock is the single most useful number on a pallet fork and is the least often stated. A fork quoted without its locking depth is a part number, not a specification.
References
- Ref 1Pallets (lever), Horopedia – Watchmaking Encyclopediasupports the description of the pallets and their locking action
- Ref 2Understanding Pallets in Lever Escapements, Fondation de la Haute Horlogeriesupports the locking, unlocking and impulse sequence described here
- Ref 3Double Roller, Horopedia – Watchmaking Encyclopediasupports the roller that passes the impulse between fork and balance
One link, one document, titled as it titles itself. No home pages, no search results.
Set this page down Set this page down when you know which of the three angles is out of spec. They fail in different ways.