Study the Certified Master Clockmaker topics by connecting theory to diagnosis: contrast count-wheel and rack-and-snail striking, trace strike counts to their root cause, apply escapement and pendulum principles to real faults, and practice restoration judgment with explicit before-and-after observations.
Telling rack-and-snail from count-wheel striking before you adjust anything
The two strike-count systems fail in different ways, so identify which system a movement uses before touching any lever. Count wheels control release by notch depth; rack-and-snail systems control count by how many rack teeth the gathering pallet collects.
A count wheel is a flat wheel, usually driven by the going train, with notches of different depths around its edge. A locking piece drops into a notch and holds the strike until the hour wheel advances it; the notch's depth corresponds to the number of blows. Faults concentrate in the lifting piece, the locking, and the wheel's registration against the hour hand.
A rack-and-snail system uses a toothed rack whose tail rests on a stepped snail driven by the hour wheel. When the strike releases, the rack falls, and a gathering pallet rakes the rack back one tooth per blow. The count is fixed by which snail step the tail landed on. Identifying the system first determines whether a wrong count points to the gathering pallet and hook, or to the wheel and its lifting pins.
- Count wheel: release governed by notch depth and locking piece; count errors implicate the wheel, its pins, or the lifting piece.
- Rack-and-snail: release governed by rack tail position on the snail; count errors implicate the rack, hook, or gathering pallet.
- Identification check: look for a rack returning after each blow versus a notched wheel advancing stepwise.
Diagnosing a wrong strike count from the sound alone
A wrong count is a symptom, not a fault. Use the sound pattern — whether the count is short, long, stuck, or off by a fixed amount — to select which part of the strike train to inspect first.
Worked scenario: a movement strikes five at seven o'clock, every hour, on a rack-and-snail system. The plausible mistake is adjusting the warning wheel or the fly because the strike 'runs badly.' The better decision is to inspect the rack tail and snail: if the tail sits on the wrong step, the count is wrong by a fixed amount every hour, which points to snail registration relative to the hour hand. Why it matters: chasing the fly wastes hours and risks damaging a train that was functioning correctly.
Contrast case: a count wheel movement repeats the previous hour's count. Here the symptom pattern differs — the count is consistently one step behind, and the deep notches reveal uneven wear. A plausible mistake is bending the lifting piece to 'add a blow,' which changes release geometry for every hour. The better decision is to check that the count wheel advances fully between hours and that the locking piece engages cleanly in each notch. The sound signature — repeat versus skip — separates wheel-side faults from lever-side faults before the plates come apart.
Strike trains: why the fly, warning, and locking must be reasoned about together
The strike train is a sequence of locked energy release: warning, then lift, then controlled run governed by the fly. Understanding that sequence turns scattered faults into a single logical path to follow.
Trace the sequence on any striking movement: the warning wheel parks against the warning pin, holding the train; at the hour, the lifting piece raises the release, the warning drops, and the train runs until the count mechanism locks it again. The fly is an air brake that controls speed, not count. This ordering explains why a clock that hangs up mid-strike usually has a locking problem, not a power problem.
Build a sound-based exercise around this sequence: with the movement in the case, listen for the pause before each strike (warning), the lift, the run, and the stop. If a clock strikes and then strikes again immediately, the locking did not re-engage after gathering — check the hook and rack teeth or the count-wheel detent. If it strikes and never stops, the stop work has failed entirely. Mapping each sound to a train stage converts an alarming symptom into a one-part inspection.
Pendulum theory you can actually apply: rate, beat error, and compensation
Rate depends on effective pendulum length, not bob weight; beat error shows as unequal tick spacing; temperature compensation manages how rod material shifts that length. Apply each principle to a distinct adjustment.
The period of a pendulum varies with the square root of its effective length, so rating nuts change rate by moving the bob relative to the suspension point, adding or removing effective length. In labeled terms: a modest rating-nut turn produces a small daily change, which is why ratings are verified over several days, not hours. Bob mass influences energy and amplitude behavior, but the timing variable is length — a distinction that decides whether you rate or look elsewhere for a fault.
Beat error is the condition where the escapement gives unequal impulse to alternate swings, audible as uneven tick-tock spacing and visible as the pendulum's being held off center. Set beat by observing the pallet engagement at the ends of swing, then confirm by listening for symmetry. Compensation is the separate problem of the rod changing length with temperature: wood, invar, and gridiron systems address it differently, and recognizing which system a clock uses tells you whether a seasonal rate change is a fault or a design characteristic.
Escapement choice: recoil versus deadbeat in context
Recoil escapements return force to the pendulum during unlock and are simple and robust; deadbeat escapements deliver impulse after lock without recoil, giving steadier rate. Each suits a different class of clock.
Anchors with recoil are found on many domestic longcase and mantel clocks: the pallets push the wheel backward slightly as the pendulum passes, which disturbs the swing but keeps the escapement forgiving of wear and variation. Their maintenance focus is entry and exit lock and drop; excessive drop wastes impulse and changes amplitude. Judging them by deadbeat standards leads to unnecessary alteration of a working escapement.
Deadbeat escapements — with pallet faces and a locking-and-impulse geometry that avoids recoil — are the regulator standard because impulse is delivered consistently and rate is steadier. Their adjustment demands precision: correct lock on both pallets, minimal drop, and clean draw. A worked check: watch the pallets at the ends of swing; in a healthy deadbeat, the tooth lands on the locking face and slides onto the impulse face with no backward motion. Observing this distinction on the bench cements why the two escapement families are maintained to different standards.
Restoration decisions: bushing versus repivoting versus leaving alone
Worn pivot holes invite three responses — fitting a bushing, turning a new pivot, or documenting and leaving a stable wear pattern. The decision should follow wear severity, originality, and the movement's role, not habit.
Worked scenario: a plate hole in a family longcase is visibly oval, but the pivot is sound. The plausible mistake is chasing the hole larger or opening it to 'clean it up,' which removes original metal and degrades future repairability. The better decision is to fit a properly sized bushing with correct alignment, preserving the plate and restoring a true bearing. Why it matters: bushing is reversible in principle, documents as a repair, and keeps the movement's history legible — exactly the judgment a master-level syllabus rewards.
Repivoting — turning a new pivot on the arbor — suits the different case where the pivot itself is bent, broken, or worn below usable size. It demands lathe work: turning, polishing, and matching the original pivot geometry. A third option, leaving well-earned wear documented and untouched, is legitimate when the wear is light and functioning is stable. The table below turns this into a decision habit rather than a reflex, and the same habit extends to hands, dials, and finishes: intervene at the level of the actual fault.
| Observation at the pivot hole | Better decision | Rationale |
|---|---|---|
| Hole oval, pivot sound and true | Fit a correctly aligned bushing | Restores bearing without sacrificing original plate metal |
| Pivot worn, scored, or bent; hole serviceable | Repivot the arbor on the lathe | Replaces the defective element instead of enlarging a sound hole |
| Light polish wear, clock runs reliably | Document and leave | Preserves originality when intervention adds no function |
| Repeated pivot breakage at one arbor | Investigate alignment and side load before repair | A new pivot in a misaligned train fails again |
Specialty clocks and a two-week practice sequence with a self-check rubric
Specialty clocks add complications — platform escapements on carriage clocks, maintaining power on regulators and skeleton clocks, multi-barrel chime sequencing — so study each as the base strike or going train plus one named addition.
Approach the specialty list systematically. Carriage clocks put the escapement on a platform, so escapement faults live outside the plates and are visible in operation. Regulators and many skeleton designs need maintaining power so the clock keeps time during winding; trace that mechanism and note what happens to the train when the mainspring is fully wound down. Chiming clocks extend the strike train: quarter chimes sequence on a separate barrel or wheel set before the hour strike releases, so a Westminster sequence fault is diagnosed one stage at a time — chime order, then sequencing between quarters, then the hour strike.
Practical exercise (bench, low risk): take any working striking clock. Week one, run it and log the strike count at each hour for three days, note the warning pause, and sketch the count system after identifying rack or count wheel. Week two, observe the pendulum's end-of-swing position for beat, then rate the clock and record daily error over four days. Self-check rubric: (1) Can you state which count system the clock uses and point to its decision part? (2) Can you describe what the warning does, in sequence? (3) Can you report whether beat was symmetric and what your daily rate showed? (4) For restoration topics, can you state, in two sentences, why bushing rather than repivoting — or the reverse — for a described fault? Score yourself per item; these are learning milestones for tracking progress, not predictions of any exam outcome.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.