Why a Watch Runs Fast After a Full Wind

Set a watch down full-wound at nine in the morning, clip the timegrapher's microphone to the caseback, and the trace will lie to you a little. Not badly — eight, maybe ten seconds fast a day, on a movement that will settle to plus two or three by evening. The first time I saw this on my own bench I assumed the regulator needed a nudge toward S. It didn't, not yet. What was happening is that the mainspring, coiled at its tightest, was pushing more torque through the train than it would an hour later, and the balance spring — the coil of blued steel actually doing the timekeeping — does not respond to that extra push with perfect indifference. Almost none of them do. The fast reading on a full wind is not really a verdict on your watch. It is a property of what a hairspring does under load, and it has a name: isochronism, or more precisely, the failure of it.

What Isochronism Is Supposed to Mean

The idea dates to Huygens and was formalized by watchmakers chasing marine chronometer accuracy: a balance spring should bring the balance wheel back to center in the same amount of time whether it swung through a wide arc or a narrow one. Big amplitude or small, the period should hold. That property, if a spring achieved it exactly, would make a watch's rate immune to changes in driving torque — full wind or nearly run down, the balance would beat at the same pace either way. In practice, no hairspring gets there. Flat balance springs, the type in the large majority of vintage and modern watches without a Breguet overcoil, are worst at it: their coils don't breathe symmetrically as they expand and contract, and the spring's effective center of gravity shifts with amplitude. That asymmetry is what shows up on the trace as a fast rate whenever amplitude runs high.

Where the Extra Torque Actually Goes

A fully wound mainspring doesn't deliver torque evenly across the whole reserve. It starts near its peak, tapers sharply at first, then eases through the middle days before falling off again near the end. That peak torque reaches the balance as extra amplitude — more energy per beat means the balance swings wider before the spring's restoring force turns it back. Once amplitude climbs past roughly 270 to 290 degrees, most balance springs start to lose their footing: the coil brushes closer to the balance cock, or simply behaves less linearly at the far end of its travel, and the beat quickens. Reading amplitude off a timegrapher trace is really the only way to confirm that's what's happening rather than guessing from the rate figure alone. It's also the same mechanism working in reverse on a chronically weak mainspring: too little torque means low amplitude, and low amplitude carries its own isochronism error, usually pulling the rate the other way.

A balance spring is not being asked to keep perfect time. It is being asked to keep the same imperfect time, twice a day, no matter how hard the mainspring happens to be pushing that morning.

Reading the Drift Across the Power Reserve

Put a full-wound watch on the timegrapher at intervals — hour one, hour twelve, day two, day five — and the rate tends to walk down from its fast start toward a settled middle, sometimes drifting slightly slow near the very end of the reserve. None of that movement is a defect on its own; it's the reason many manufacturers spec a caliber's rate as an average across the reserve rather than a single number taken fresh off the winder. The table below is representative of what a movement in reasonably good adjustment tends to show. Figures vary by caliber, spring alloy, and whether the spring carries an overcoil.

Isochronism vs. an Actual Regulation Fault

The trouble is that isochronism error and a genuine fault look similar on a single reading — both show up as “the watch runs fast.” The difference is in the shape of the drift over time, not the number on any one trace. A spring simply behaving as springs do will ease back toward center as the day goes on and stay roughly consistent from one wind to the next. A real fault tends to do one of these instead:

• Rate that stays fast well into the second day, past when torque should have settled — points toward a bent or distorted hairspring holding a wider arc than it should at any torque level.
• Amplitude that's unusually high across the entire reserve, not just at full wind — worth checking the mainspring for excessive set, or the barrel for a slipping bridle.
• A fast reading that swings wildly by position (dial up versus crown down) well beyond the normal spread — that's positional error compounding on top of isochronism, and it deserves its own look rather than being blamed on the wind alone.
• A fast drift that repeats identically at every single wind with no settling at all — sometimes traced to a magnetized spring, which is a torque-independent problem wearing isochronism's clothes.

Living With a Few Seconds

None of this means a fast reading at full wind should be waved off outright. A movement that's ten or fifteen seconds fast on day one and never settles has told you something worth chasing further. But a movement that reads +8 at nine in the morning and +2 by evening, day after day, isn't broken. It's a spring doing what flat balance springs have always done under load. A good part of learning to regulate at a kitchen table is learning which reading in the day actually represents the watch — usually the settled one, taken twelve to twenty-four hours after a full wind, not the first trace off a fresh one.