A watch stop is a small lever-and-spring mechanism inside a mechanical watch that physically halts the balance wheel when you pull the crown to the time-setting position. The Omega Speedmaster Professional 1861 and the ETA 2824-2 both use one. Pulling the stem trips a hacking lever that drops a thin spring against the balance rim, freezing the seconds hand so you can synchronise the watch to a reference signal down to the second. Release the crown and the balance restarts within 1-2 oscillations.
Watch Stop Interactive Calculator
Vary beat rate, balance amplitude, contact-force band, and restart oscillations to see the hacking lever timing and force window.
Equation Used
The calculator converts watch beat rate from vibrations per hour to full balance oscillations per second. The period of one oscillation sets the estimated restart time for the 1-2 oscillation recovery described in the article, while the stop-spring force inputs define the acceptable contact-force band.
- VPH is watch vibrations per hour, with two vibrations per full balance oscillation.
- Restart time is estimated from 1 to 2 full balance oscillations after release.
- Contact force band is treated as the allowable stop-spring range from the article.
- Amplitude is used for visualization of balance motion, not for stored energy calculation.
Inside the Watch Stop
The watch stop — watchmakers call it a hacking lever or hack feature — sits on the train side of the movement and connects to the stem through the keyless works. When you pull the crown out to position 2 (the time-setting position), the setting lever rocks, the yoke shifts, and a small tail on the hacking lever rides up a cam or flag on the setting lever. That motion swings the free end of the hacking lever — usually a thin flat spring or a hardened steel finger — into the rim of the balance wheel. The balance stops dead. The seconds hand, geared directly to the fourth wheel which the balance controls through the escapement, freezes with it.
Geometry matters here more than people realise. The stop spring must contact the balance rim with enough force to overcome the residual amplitude of the oscillating balance, but light enough that it does not bend the rim, mark the polished edge, or push the balance staff against its jewel. On a 28,800 vph movement like the ETA 2824-2, the balance still has roughly 270° of amplitude when you hack it, so the spring sees a moving target with real momentum. Contact force in the 0.5-2 mN range is typical. Too soft and the balance keeps creeping. Too stiff and you damage the rim or — worse — the spring stays sprung against the rim after you push the crown back in, and the watch will not start.
The most common failure mode you will see on the bench is a hacking lever that no longer fully retracts. Causes are predictable: a weak return spring, dried lubricant on the lever pivot, a bent stop finger, or wear on the cam surface where the lever tail rides. If the watch runs only when the crown is pushed firmly home and stops when the crown is at rest in position 1, suspect the hacking lever first.
Key Components
- Hacking Lever (Stop Lever): The pivoted lever that carries the stop spring. Typically 0.20-0.25 mm thick hardened steel, pivoted on a single screw or post on the train-side bridge. Its tail rides on the setting lever or yoke; its working end carries the spring that contacts the balance rim.
- Stop Spring: The thin flexing finger — sometimes integral to the hacking lever, sometimes a separate riveted leaf — that touches the balance rim. Contact force must sit in the 0.5-2 mN window. Too high and the rim marks; too low and the balance creeps.
- Setting Lever Cam: A small flag, ramp, or step on the setting lever that lifts the hacking lever tail when the stem is pulled to position 2. Cam profile must give clean engagement within the first 0.3-0.5 mm of stem travel so the watch hacks before the user feels the click into the time-setting detent.
- Return Spring: A separate light spring (or a spring tail formed into the hacking lever itself) that pulls the lever clear of the balance when the crown returns to position 1. Spring force is the only thing standing between you and a watch that will not start, so it must dominate any pivot friction by at least 3:1.
- Balance Wheel Rim: The contact surface for the stop spring. On a Glucydur or beryllium-copper balance the rim is polished and 0.4-0.6 mm wide. The spring must land on the rim, never on the screws or arms, or you risk knocking the poise out and changing rate by 5-15 seconds per day.
Who Uses the Watch Stop
Hacking is not standard on every mechanical watch — plenty of vintage calibres, including the Rolex 1570 series before 1972, ran without it. The stop feature appears wherever the user needs to set the time precisely against an external reference, which means military watches, chronometers, dive watches, and any movement marketed as a precision instrument. Modern Swiss base calibres added it almost universally by the 1990s because consumer expectation shifted — buyers expect to synchronise to a phone or radio signal.
- Mechanical Wristwatches: ETA 2824-2 and its Sellita SW200 clone — the workhorse three-hand automatic in thousands of microbrand watches, hacking added as standard from the 2824 generation onward.
- Military & Field Watches: MIL-W-46374 specification field watches and the Hamilton Khaki Field Mechanical (calibre H-50) — hacking was a military requirement for unit-wide time synchronisation before patrol.
- Dive Watches: The Tudor Black Bay calibre MT5602 hacks so the diver can sync to dive-computer time at descent.
- Chronographs: The Omega Speedmaster Professional calibre 1861 hacks the running seconds for accurate event timing reference, even though the chronograph subdial does not.
- Marine Chronometers & Deck Watches: Hamilton Model 22 deck watches used a hacking feature so navigators could rate the watch against a chronometer or radio time signal during celestial observation.
- Pilot Watches: IWC Mark XI, supplied to the RAF from 1948, used a hacking calibre 89 for flight-deck time synchronisation.
The Formula Behind the Watch Stop
The useful number to compute is the contact force the stop spring applies to the balance rim. Get this wrong in either direction and the mechanism fails. At the low end of the typical force range — around 0.5 mN — a healthy 270° amplitude balance can sometimes ride through the contact and creep slowly forward, especially if the spring lands at a shallow tangent. At the high end, around 2 mN, you reliably stop the balance but you also start to risk marking the rim and pushing the staff laterally against its jewel. The sweet spot for a standard 9-10 mm balance running at 28,800 vph sits around 1.0-1.5 mN. The formula treats the stop spring as a cantilever beam and solves for tip force from measured deflection.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| Fstop | Contact force applied by stop spring tip to balance rim | N | lbf |
| E | Young's modulus of spring steel | Pa | psi |
| I | Second moment of area of spring cross-section, b·h3/12 | m4 | in4 |
| δ | Deflection of spring tip when engaged against balance rim | m | in |
| L | Free length of stop spring from clamp to contact tip | m | in |
Worked Example: Watch Stop in an ETA 2824-2 service rebuild
A watchmaker in Bienne is rebuilding an ETA 2824-2 movement and replacing a damaged stop spring. The replacement is a flat hardened steel leaf, 0.15 mm thick, 0.50 mm wide, with a free length of 4.0 mm from its rivet to the contact tip. When the crown pulls to position 2, the spring tip deflects 0.10 mm against the balance rim. He needs to confirm the contact force sits inside the 0.5-2 mN window before reassembly.
Given
- E = 200 × 109 Pa
- b (width) = 0.50 mm
- h (thickness) = 0.15 mm
- L = 4.0 mm
- δ (nominal) = 0.10 mm
Solution
Step 1 — compute the second moment of area for the rectangular spring cross-section:
Step 2 — at the nominal deflection of 0.10 mm, solve for tip force:
That sits squarely in the design sweet spot. The balance halts cleanly within one or two oscillations and the rim sees no marking risk.
Step 3 — at the low end of likely deflection variation, δ = 0.05 mm (you bent the spring tip too far back during fitting):
That is at the bottom of the working window. The balance still stops, but on a watch with strong amplitude — fresh service, fully wound, dial up — you may see the seconds hand creep forward by half a tooth before locking. Some users will complain the watch does not hack reliably.
Step 4 — at the high end, δ = 0.20 mm (spring set too aggressively, or a bent contact tip):
Now you are above the safe range. The balance stops instantly, but you risk marking the polished rim and pushing the balance staff sideways against the lower jewel, which can shift poise and add 5-15 seconds per day to the rate after you let the crown back in.
Result
The nominal stop spring delivers 1. 3 mN of contact force — right in the centre of the design window for a 28,800 vph ETA 2824-2. At 0.66 mN (low-deflection case) the watch will still hack but the seconds hand may creep slightly before locking; at 2.64 mN (high-deflection case) you stop the balance hard but risk marking the rim and shifting balance poise. If your measured force or measured rim contact disagrees with the prediction, the three usual culprits are: (1) the stop spring tip is hitting an arm or screw of the balance rather than the rim itself — check engagement under a loupe at 10x and reshape the tip if needed; (2) the spring is contacting at a steep angle rather than tangentially, which converts contact force into a sideways push on the balance staff; (3) the spring length L was measured from the wrong datum — L is from the clamp/rivet line, not from the lever pivot, and a 0.5 mm error in L changes the force by roughly a factor of 2 because the term is cubed.
Watch Stop vs Alternatives
Hacking is not free. It adds parts, adds friction in the keyless works, and gives the watchmaker one more thing to adjust at service. The decision is whether the precision benefit justifies the cost and complexity, and whether you really need it on this calibre. Compare it against the two real alternatives: a non-hacking movement, and a quartz movement (which by its nature stops the second hand whenever you pull the crown).
| Property | Watch Stop (mechanical hack) | Non-hacking mechanical | Quartz with stop-second |
|---|---|---|---|
| Setting precision (synchronise to reference) | ±0.5 second | ±5-10 seconds (estimate the swing) | ±0.1 second |
| Added parts vs base calibre | 1 lever + 1 spring (sometimes integral) | 0 | 0 — handled by IC firmware |
| Restart time after crown push-in | 1-2 oscillations (~0.5 sec) | Immediate (never stopped) | Immediate from quartz timebase |
| Failure mode if it goes wrong | Watch refuses to start (lever stuck against rim) | None — feature absent | Pull-out switch contact failure |
| Service complexity | Lever pivot lubrication, stop spring inspection at each service | Nothing extra | Nothing extra |
| Typical application | Modern Swiss base calibres, military, dive, chronometer | Vintage calibres, dress watches where set-precision is unimportant | All quartz movements as a default feature |
Frequently Asked Questions About Watch Stop
The hacking lever is not retracting fully. The stop spring is staying in light contact with the rim even with the crown home, and that residual drag is enough to kill a balance that has not built up amplitude yet.
Three causes to check, in order of likelihood. First, the return spring on the hacking lever is too weak or has been replaced with the wrong part — it must overcome pivot friction by at least 3:1. Second, the lever pivot is dry or has dried oil that has gone gummy; clean and lubricate with a light oil like Moebius 9010, sparingly. Third, the cam profile on the setting lever is worn, so the lever tail does not fall fully off the cam when the stem returns. Look for a polished step where the tail rides — if you see one, the setting lever needs replacement.
Pull the crown to position 2 with the movement on a microscope or 10x loupe and watch the engagement. The spring tip should land on the flat polished outer band of the rim, between the timing screws, with the contact tangential — meaning the spring approaches the rim along its length, not stabbing into it perpendicular.
If you see the tip hitting a balance screw or one of the spokes, the spring is bent out of plane or the lever pivot has shifted. If contact is perpendicular rather than tangential you will hear a faint click each time you hack the watch, and the rate will drift after a few hundred hack cycles because you are slowly knocking the balance out of poise.
You are almost certainly knocking the balance out of poise. Every time the stop spring lands on the rim, it applies a small impulse. If the contact force is at the high end (above ~2 mN) or if the spring is hitting at an angle rather than tangentially, that impulse slowly displaces material — or, more commonly, slightly bends the rim or shifts a timing screw.
Put the watch on a timing machine in six positions before and after a week of frequent hacking. If positional variation has increased — particularly between dial-up and crown-down — poise has shifted. The fix is to measure the actual stop spring force using the cantilever formula in this article, reshape or replace the spring if it is over-stressed, and re-poise the balance.
No. The temptation exists but the engineering case is bad. These movements were designed with their bridges, train layout, and keyless works finalised without space for a hacking lever. Aftermarket hack mods exist for some calibres but they typically work by adding a friction brake against the seconds wheel pivot rather than the balance rim, which adds wear to a pivot that was never designed for side load.
If precise time-setting matters to you, buy a calibre that was designed with a hack from the start — the 1575 replaced the 1570 in 1971 specifically to add hacking. Do not modify a vintage movement; you reduce its value and add a failure point.
The cantilever formula assumes the spring is anchored rigidly and contacts the rim at the calculated free-length tip. In practice, two things commonly violate that assumption.
First, the hacking lever itself has compliance. If the lever is thin or the pivot is loose, some of your spring deflection is being absorbed by lever bending rather than spring bending, so the actual spring force at the rim is lower than predicted. Second, the rim contact may be sliding rather than gripping. A rim with residual oil contamination or a polished spring tip with no friction texture will let the rim creep under steady force. Degrease the rim with a fine brush and naphtha, and check that the spring tip has a slight matte finish where it contacts.
It should not, and if it does you have a fault. With the crown at position 1, the stop spring should sit at least 0.1-0.2 mm clear of the balance rim. There should be visible daylight under a loupe. If you measure amplitude drop of 10° or more compared to the same calibre without a hack, the spring is dragging.
A useful diagnostic: run the watch on a timing machine, then carefully lift the hacking lever clear with a fine peg. If amplitude jumps and rate changes, the spring was in residual contact. The fix is usually to reform the spring's rest geometry so it sits further from the rim, or to strengthen the return spring on the lever.
References & Further Reading
- Wikipedia contributors. Hack watch. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
