Watch Beat Error Calculator for US Watchmakers and Timegrapher Readings

The correct sequence for regulating any mechanical movement is: beat error first, then rate. Adjusting the regulator arm changes rate but not beat error; adjusting the beat setting changes beat centering without directly altering rate, though it may shift the rate reading slightly as the escapement geometry changes. Starting with the regulator when beat error is high means you are chasing a moving target. The rate reading on a movement with 1.5 ms beat error is less meaningful than the same reading at 0.2 ms, because the asymmetric oscillation means positional rate variation will dominate the apparent rate number. Get the beat error under 0.5 ms first, then address rate by moving the regulator, then re-check beat error to confirm the regulator movement did not disturb the beat setting lever.

Beat error correction on different movement types requires different tools and techniques. Movements with a dedicated beat setting lever (most modern Swiss and Japanese auto movements including ETA 2824, SW200, NH35/36, Miyota 82-series) allow in-case adjustment by moving the lever one click at a time with a fine oiler or dedicated beat setting tool, without removing the balance. Movements with an eccentric cam on the collet (found on some vintage movements including Hamilton, Waltham railroad grades) require more careful incremental adjustment. Movements with a fixed collet (traditional construction) require removing the balance, repositioning the impulse jewel on the collet using a hand staking set, and re-installing the balance before testing. In all cases, make adjustments in increments of roughly 0.5 to 1 degree and re-test on the timegrapher after each increment. Never try to correct 2+ ms of beat error in a single large adjustment.

Timegrapher testing should always begin with the movement in dial-up (horizontal) position, as this minimizes gravitational effects on the balance and gives the most consistent baseline reading. After achieving acceptable beat error in dial-up, always check in at least one vertical position (crown-up is most revealing) to confirm positional variation is within normal range. A beat error that looks good in dial-up but explodes to 2.0+ ms in crown-up is a diagnostic indicator, not a beat setting problem. It points to balance poising issues (uneven mass distribution on the balance wheel), worn pivots causing the balance to run off-center under gravity, or a bent balance staff. No amount of beat setting adjustment will resolve positional beat error variation caused by mechanical damage. Five-position testing, as supported by this calculator’s tracking mode, gives the most complete diagnostic picture for a movement that will be worn in multiple orientations throughout the day.

One of the most commonly made errors by hobbyist watchmakers using a timegrapher for the first time is believing that the lift angle setting affects the beat error reading. It does not. As Escapement.watch noted in their April 2026 lift angle guide, the lift angle setting only affects the amplitude calculation, because amplitude is derived from the time between the first and third acoustic pulses of the escapement signal (which are separated by the lift angle of the escape wheel tooth). The beat error is calculated from the time difference between corresponding pulses in consecutive beats, which is entirely independent of lift angle. This means you can have an incorrect lift angle setting and still get an accurate beat error reading, but your amplitude number will be wrong. Always verify the lift angle for your specific caliber before trusting the amplitude reading. Common values: 50° for ETA 2824 and SW200; 49° for Miyota 82-series; 53° for Seiko NH35/36; 38° for Omega Co-Axial calibers. The caliber database in this calculator pre-loads the correct lift angle for reference alongside the BPH.

The Witschi WAIO thresholds of 0.5 ms excellent and 0.8 ms upper tolerance are calibrated for modern lever escapement wristwatches running at 21,600 to 28,800 BPH. Vintage American pocket watch movements, particularly the 18-jewel and 21-jewel railroad-grade movements from Hamilton (992B, 950B), Waltham (Model 1908), Illinois (Bunn Special), and Ball Railroad Standard that operated at 18,000 BPH, have a longer half-period (100 ms at 18,000 BPH versus 62.5 ms at 28,800 BPH). The same degree of impulse jewel offset produces a smaller millisecond beat error at 18,000 BPH than at 28,800 BPH. AWCI training materials and standard vintage watchmaking practice therefore apply more lenient thresholds: beat error up to 0.8 ms is considered acceptable for vintage pocket watch movements in good mechanical condition, and beat error up to 1.5 ms warrants attention but does not necessarily prevent reliable operation. This calculator’s vintage mode applies these appropriate thresholds when you select the vintage movement type, so the assessment zones properly reflect the different standards for these historically important American timepieces. For a deep reference on railroad-grade movements, the NAWCC’s Martin Luther Archival Library at Columbia, Pennsylvania, maintains the most complete collection of American pocket watch service documentation in the United States.

Professional watchmakers at AWCI-certified service centers document timegrapher readings at every stage of a service: pre-service (as-received condition), post-service before case-up (movement alone on the timegrapher), and post-service after casing (which can change readings due to case pressure on the movement). This documentation is not bureaucratic overhead; it is quality assurance, diagnostic intelligence, and professional protection. A customer who returns a week later claiming the watch stopped can be shown the pre-service reading (perhaps 2.5 ms beat error and 180° amplitude at low reserve) versus the post-service reading (0.3 ms and 285°), demonstrating that the service was performed correctly and the current issue is unrelated. This calculator’s PDF report generates a complete bench documentation record with caliber details, all readings, assessment zones, formula references, and data sources, suitable for inclusion in a service file or customer communication. The AWCI’s professional practice guidelines recommend maintaining service records for a minimum of seven years.