Diagnosing a Non-Runner: The Four-Step Process

A movement lands on your bench and doesn’t run. Or it runs, but so poorly you’d call it the same thing. The usual first move is to start poking around hoping something jumps out. Sometimes it does. More often, thirty minutes later you’re looking at a partially disassembled movement with no clearer picture of the problem than when you started.

There’s a better approach, and it works on any non-runner.

The framework has four steps: Condition, Cause, Correction, and Confirm. The condition is what you’re dealing with, in this case a movement that won’t run or runs so badly it might as well not. The cause is what you find through systematic isolation. The correction is the fix. The confirm step is where you verify the fix actually solved the problem and didn’t introduce a new one.

Since roughly ninety percent of non-runner problems originate in the balance or escapement area, that’s where the process starts.

Visual Inspection

Before you touch anything, look.

Take an artist’s brush and lightly sweep it back and forth across the balance rim. What you’re checking is whether the balance moves vertically or laterally in response. Either would suggest a broken pivot. If a pivot appears broken, remove the balance wheel and confirm it before going further.

If the pivots look intact, check the hairspring. It should be centered on its collet and appear perpendicular to the balance wheel. While you’re there, check that the balance rim itself is perpendicular to the collet. If it looks skewed to one side, note it.

Next, check the balance wheel’s endshake. This is the vertical movement of the balance arbor between its pivots, and the correct amount is in the hundredths of a millimeter, so you’re not measuring it directly. What you’re doing is comparing it visually to the endshake of the pallet fork and the escape wheel. All three should appear roughly equal. If the balance endshake looks noticeably greater than the other two, that’s a problem worth investigating.

The Rotation Test

With the initial visual check done, test how freely the balance actually rotates.

Use a brush or a piece of pegwood to bring the balance rim around to 180 degrees and release it. Count how long it rotates before stopping. Do this in both horizontal positions, dial up and dial down. The timing should be nearly identical in both positions, because the forces acting on the balance (pivot friction, the hairspring’s response to gravity) are the same from either side. If one position runs noticeably shorter than the other, the problem is most likely located under the balance wheel in the worse position.

While the balance is rotating, watch how it moves. It should stay on a consistent plane as it oscillates, not bob up and down. Watch how it stops. A smooth, gradual stop is normal. A sudden, jerky stop can indicate a bent pivot creating friction inside the balance jewel.

Three outcomes are possible. The balance runs poorly in both positions. It runs better in one than the other. Or it doesn’t run at all. If it doesn’t run at all, the problem could be in the balance itself or in the escapement, and at this point you’re still collecting information, not drawing conclusions.

One thing worth checking now, particularly if the movement was recently worked on: the impulse jewel may be sitting outside the pallet fork slot rather than passing through it. This prevents the balance from oscillating entirely and is a common assembly mistake for anyone learning. If that’s the case, remove the balance wheel and reinstall it with the impulse jewel correctly seated in the fork slot.

Pre-Cleaning

At this stage, everything is dirty, and grit or dried lubricant can affect how the balance behaves in ways that have nothing to do with underlying mechanical problems. Before drawing any conclusions from the rotation test, pre-clean the parts and run the test again.

Remove the balance wheel and pallet fork. Remove the balance jewels and clean them by whatever method you use. Give the balance wheel a brief soak in heptane. While you’re at it, do a close visual inspection of the balance pivots under magnification. You’re looking for anything that obviously looks wrong.

If you find a bent pivot, source a balance complete. A balance complete is a hairspring and balance wheel that have been matched at the factory. Watchmakers with more experience will replace just the balance staff instead. Don’t attempt to straighten a bent pivot. Once a pivot is bent, there’s no way to restore it accurately enough for the watch to keep reliable time. That’s not a technique limitation, it’s a geometry problem.

If the pivots look straight, check for corrosion. Light corrosion can sometimes be addressed with a pin polisher followed by a re-clean with heptane. Heavy corrosion, pitting, or visible scratches on the pivot surface means the staff needs replacement. A pitted pivot will never run cleanly against a jewel.

While the jewels are out, inspect them carefully. Cracks in a balance jewel are easy to miss but they do affect running, particularly in the vertical positions. Check the endstones as well. Pitting or wear marks on the endstone surface cause the same kind of friction problem.

Check the impulse jewel. It should sit at 90 degrees to the roller table and should not be tilted left or right when you look at the roller table straight on. An impulse jewel that’s off-angle will prevent the watch from running, or at the very least prevent it from running well, depending on the direction and degree of the tilt.

Finally, look at the hairspring with the balance flipped upside down and the upper pivot seated in its jewel hole. The coils should be centered on the collet, the spacing between coils should be even, and no coils should be touching. Hairspring repair takes time to develop properly. If this movement is important to you or it’s valuable, don’t attempt a hairspring correction on it. Buy a supply of miscellaneous hairsprings and practice on those first. The skill has to be built before it can be trusted.

Isolation Test

Once the pivots are clean and everything checks out visually, reinstall the lubricated balance jewels and put the balance back into the movement, but leave the pallet fork out. Testing the balance without the pallet fork removes the escapement from the equation entirely and isolates whether the problem is in the balance or somewhere in the escapement.

Repeat the 180-degree rotation test in both horizontal positions. With clean pivots and a balance in good condition, you should see about 30 seconds of rotation in each position before a smooth stop.

If one position still runs shorter than the other, revisit what’s under the balance wheel in that position. On movements with shock-protected jewels, check whether the endstones were accidentally swapped during cleaning. They’re not always the same size, and running the wrong endstone in the wrong position is enough to affect endshake and cause this. On movements where the jewel settings are held in by screws, make sure the settings are sitting flush and the screws are snug. A loose or cocked setting can cause the pivot to ride on the conical part of the jewel rather than seating correctly, and the balance will run slower in that position as a result.

If the balance is now running correctly, either the pre-cleaning removed something that was preventing it from oscillating, you replaced a broken jewel, or you replaced the balance staff. In any case, the balance is working. Now it needs to be confirmed with the escapement back in the picture.

The Pallet Fork

There’s a reason we haven’t looked at the pallet fork until now.

Non-runner problems often have more than one cause. Solving them one system at a time means you know exactly what each fix contributed. If you inspect the balance and the pallet fork simultaneously and correct both, you don’t know which fix solved the problem or whether one correction created something new. Fix the balance first. Confirm it’s working. Then look at the pallet fork.

The pallet fork transfers power from the mainspring to the balance through impulse. When it fails to do that, the balance either stops or runs poorly.

Check each pallet stone first. If the face of a stone is chipped or damaged, an escape wheel tooth can slip past it rather than locking cleanly, and the transfer of energy breaks down. Push lightly on each stone with a sharpened piece of pegwood or the tip of a large oiler. The stone should be completely solid. Any movement means the shellac holding it in the fork slot has failed. This is more common in movements that haven’t been properly cleaned, and it’s worth checking even when everything else looks fine.

Check the lever of the pallet fork. A bent lever puts the escapement out of angle and can cause the watch to run poorly or stop entirely.

Check the guard pin. It should be straight in its slot at the end of the pallet fork. A bent guard pin drags across the roller table and acts as a brake on the balance wheel.

Check the pallet fork pivots for any obvious damage.

Escape Wheel Check

Before reinstalling the balance, cycle the pallet fork back and forth while watching the escape wheel teeth land on the pallet stones. Work through all fifteen teeth on both the entry and exit sides. You’re looking for consistent landing position on each stone and checking the condition of the teeth themselves. A rounded or bent escape wheel tooth will cause problems that no amount of balance work will fix.

This isn’t a complete escapement inspection, and it won’t catch geometry problems that require more advanced techniques. But it will catch obvious mechanical damage before you close everything up.

Confirming the Repair

With the pallet fork cleaned and reinstalled, repeat the rotation test one more time.

Bring the balance around to 180 degrees and release it. Time both horizontal positions. If the balance runs as well with the pallet fork installed as it did without it, the problem was in the balance and you’ve already corrected it.

If the balance was running correctly without the pallet fork and is now running poorly, you’ve isolated the problem to the pallet fork itself. For a watchmaker who’s still learning, the most straightforward option is to source a replacement. One caution on older movements with adjustable banking pins: a replacement fork will usually need adjustment before the escapement works properly. The manufacturing tolerances on movements from eighty or a hundred years ago are nothing like what’s produced today, even in low-end movements. A direct swap rarely works without some additional escapement work.

That’s the process: condition, cause, correction, confirm. It won’t resolve every non-runner, and some problems, particularly advanced escapement geometry, require techniques that go well beyond this walk-through. But for most of what comes across the bench, working through this sequence will isolate the problem without having to take the movement apart twice.