The Duplex Escapement is a single-beat horological escapement that uses an escape wheel with two sets of teeth — long locking teeth on the wheel rim and short impulse teeth on the wheel face — to control a balance wheel. Unlike the lever escapement that delivers impulse on every swing, the Duplex Escapement only impulses on alternating swings, halving frictional disturbance to the balance. It existed to give better rate accuracy than the verge while staying cheaper than a chronometer, and it powered hundreds of thousands of high-grade pocket watches between roughly 1780 and 1860.
Duplex Escapement Interactive Calculator
Vary balance rate, tooth count, and single-beat timing to see the escape wheel impulse rate and seconds-hand advance.
Equation Used
The duplex escapement is single-beat: the escape wheel receives one impulse after a set number of balance swings. The calculator converts balance beats per hour into impulses per second, then divides by escape-wheel tooth count to estimate the seconds-hand wheel speed and daily rate error.
- One escape-wheel tooth advances per impulse.
- Duplex action impulses once every selected number of balance swings.
- Seconds hand is driven from the selected escape-wheel rate.
- Ideal timing: no tripping, slip, or banking loss.
The Duplex Escapement in Action
The Duplex Escapement, also known as the Guernsey Escapement in some 19th-century English trade catalogues, locks the gear train against a long tooth on the escape wheel rim using a small ruby roller fitted to the balance staff. As the balance swings in the unlocking direction, a slot cut into the ruby roller passes the locking tooth — the wheel jumps forward by a fraction of a tooth, and a short impulse tooth standing up from the wheel face strikes a separate pallet on the balance staff to deliver energy. On the return swing, the locking tooth simply slides over the curved face of the ruby roller without unlocking. So the balance receives one impulse for every two swings — that's why horologists call this a single-beat escapement.
The geometry is unforgiving. The locking-tooth tip clearance against the ruby roller must sit around 0.02 to 0.04 mm — too tight and the wheel jams on temperature swing, too loose and the wheel "trips," meaning a tooth slips past during a small balance excursion and the watch gallops forward by half a turn of the escape wheel. If you notice the watch losing time only when worn (not on the bench), suspect tripping caused by shock-induced over-banking. The impulse pallet must be set so its strike face meets the impulse tooth at the precise instant the roller slot has cleared the locking tooth — miss that timing window by more than about 1° of balance rotation and the watch either won't self-start or will stop on the slightest knock.
What makes the Duplex Escapement (watches version specifically, as opposed to the rare clock variant) attractive is the friction-rest behaviour. While the wheel is locked against the roller, the only disturbance to the balance is sliding friction across a polished ruby surface — extremely low. That low disturbance is why a well-made duplex pocket watch can hold rate to within 5 seconds per day, rivalling a lever watch of the same era despite being mechanically simpler.
Key Components
- Escape wheel (duplex form): A single wheel carrying two tooth sets — typically 15 long locking teeth around the rim and 15 short impulse teeth standing axially from the wheel face. The two tooth sets must be angularly aligned within about 0.5° or the impulse arrives at the wrong instant relative to unlocking.
- Ruby roller (locking roller): A small cylindrical ruby fitted on the balance staff with a narrow radial slot machined through it. The slot width controls unlocking timing — typically 0.15 to 0.25 mm wide for a 12-line watch — and the cylindrical face holds the escape wheel locked during the dead swing.
- Impulse pallet: A separate small ruby or steel finger projecting from the balance staff above the roller. It catches the short impulse tooth once per pair of balance swings and converts the wheel's stored energy into balance amplitude.
- Balance wheel and hairspring: Standard bimetallic or monometallic balance running at 14,400 or 18,000 bph. Because impulse only arrives every other swing, the balance must carry enough inertia to coast through the dead swing without dropping below the unlocking amplitude — typically a minimum of 180° of arc.
- Banking pins: Limit balance overswing to prevent the impulse pallet from re-entering the wheel from the wrong side. Set with about 5° of safety margin past the unlocking point — too tight kills amplitude, too loose allows the pallet to clip the next impulse tooth and stop the watch.
Real-World Applications of the Duplex Escapement
The Duplex Escapement found its commercial sweet spot in mid-grade pocket watches where buyers wanted better timekeeping than a verge could offer but couldn't justify the cost of a detent chronometer or a fully jewelled lever. It dominated several specific market segments for about 80 years and still appears in restoration work today.
- English pocket watch manufacturing: Late-18th and early-19th century London makers including Thomas Tyrer (who patented an early form in 1782) and Barwise produced thousands of duplex pocket watches for the carriage trade. The Guernsey Escapement name attached to a Channel Islands manufacturing variant sold through English chandlers.
- Chinese market export watches: Swiss and English makers including Bovet and Ilbery produced ornate duplex pocket watches specifically for export to Qing-dynasty China between 1820 and 1860 — often with painted enamel cases and centre-seconds dead-beat seconds hands driven directly by the once-per-two-beat impulse.
- American mass-production watchmaking: The Waterbury Watch Company built the so-called "long-wind" duplex pocket watch in Connecticut from 1880 onward — a low-cost duplex driven by an extremely long mainspring as a deliberately cheap alternative to a lever movement.
- Antique horology restoration: Specialist restorers working on Bovet, Ilbery, and Tyrer pieces still cut replacement ruby rollers and impulse pallets to original geometry — the parts are not interchangeable between makers because each used slightly different tooth pitch and roller diameter.
- Horological education: Programmes such as WOSTEP in Neuchâtel and BHI in Upton Hall use cutaway duplex movements to teach single-beat escapement theory before students move on to detent chronometer work.
- Museum demonstration timepieces: The British Museum and the Patek Philippe Museum in Geneva display working oversize duplex demonstrators that let visitors see the locking-tooth and impulse-tooth interaction at hand-cranked speed.
The Formula Behind the Duplex Escapement
The most useful design calculation for a Duplex Escapement is the relationship between balance frequency, escape wheel tooth count, and the rate at which the seconds hand advances. Because impulse occurs only every other swing, the seconds hand of a duplex watch jumps once per second on a 18,000 bph movement — not five times per second like a modern lever watch. At the low end of practical balance frequency (12,600 bph) the seconds hand crawls at 0.7 ticks per second and the watch feels sluggish; at 18,000 bph (nominal for this escapement) you get the characteristic clean once-per-second tick that buyers in 1820 prized as a chronometer-like feature; pushing toward 21,600 bph drives the locking-tooth impact velocity high enough that ruby rollers begin to chip within months of service. The sweet spot for a duplex is firmly at 18,000 bph with a 15-tooth wheel.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| ftick | Seconds-hand tick frequency (audible/visible advances per second) | Hz | ticks/sec |
| bph | Balance beats per hour (one beat = one swing in either direction) | beats/hr | beats/hr |
| Z | Number of escape wheel teeth (locking-tooth count) | teeth | teeth |
| 7200 | Constant: 3600 sec/hr × 2 (single-beat escapement impulses every other swing) | — | — |
Worked Example: Duplex Escapement in an 1830s Bovet Chinese-market duplex pocket watch
A specialist restorer in Edinburgh receives an 1830s Bovet duplex pocket watch built for the Canton trade. The escape wheel carries 15 locking teeth, the balance is signed for 18,000 bph, and the centre-seconds hand should advance once per second exactly. The customer reports the seconds hand sometimes jumps two divisions instead of one. You need to verify the design tick rate, work out what the watch does at 12,600 bph and 21,600 bph for comparison, and decide whether the symptom points to a tripping fault.
Given
- bphnom = 18000 beats/hr
- Z = 15 teeth
- bphlow = 12600 beats/hr
- bphhigh = 21600 beats/hr
Solution
Step 1 — at the nominal 18,000 bph design point, calculate the seconds-hand tick rate using the formula:
The formula reduces correctly only when the centre-seconds gearing matches the escape wheel rotation rate. For this Bovet layout the design intent is ftick,nom = 1.0 Hz — one clean jump per second, which is exactly what made these watches commercially desirable in 1830.
Step 2 — at the low end of plausible balance frequency (12,600 bph, what you'd see if the hairspring has been mis-vibrated or the balance is running at low amplitude):
The seconds hand would crawl visibly slow — a customer would notice the watch losing roughly 7 hours per day. This is so far off design that it indicates a hairspring problem, not normal operation.
Step 3 — at the high end (21,600 bph, what you'd see if someone has fitted a stiffer hairspring trying to "improve" the watch):
The seconds hand ticks 20% fast — about 4 hours per day gain. More dangerously, the locking-tooth impact velocity at the ruby roller scales with the square of balance frequency, so impact energy is up by 44%. Ruby rollers historically chip within 3 to 6 months at this rate. A duplex is firmly designed for 18,000 bph and nothing else.
Result
Nominal design rate is 1. 0 Hz — one seconds-hand tick per second exactly, which is what the customer should be seeing. At the low operating point of 12,600 bph the hand would crawl at 0.7 Hz losing 7 hours a day; at the high point of 21,600 bph it ticks at 1.2 Hz gaining 4 hours a day. The customer's reported symptom — the hand sometimes jumping two divisions instead of one — does not match either rate error; it points to escapement tripping, where shock causes the locking tooth to slip past the ruby roller during a small balance excursion. Most likely causes: (1) ruby roller slot edge chipped or worn so the unlocking geometry is sloppy, (2) impulse pallet set too deep so it clips the next impulse tooth on a recoil and drives the wheel forward an extra step, or (3) balance amplitude dropping below 180° on the dial-up position, letting the wheel unlock on what should be a dead swing. Pull the balance and inspect the ruby roller slot under 40× magnification — chipping there is the single most common cause of duplex tripping in surviving 1830s pieces.
When to Use a Duplex Escapement and When Not To
The Duplex Escapement competed directly with the verge, the cylinder, the lever, and the detent chronometer in the 1780-1860 pocket watch market. Each had a different cost, accuracy, and shock-tolerance profile, and the duplex sat in a specific commercial niche between the cheap cylinder and the expensive chronometer. The comparison below uses the engineering dimensions a restorer or a movement designer actually cares about.
| Property | Duplex Escapement | Lever Escapement | Detent Chronometer Escapement |
|---|---|---|---|
| Typical rate accuracy (well-made, period example) | ±5 sec/day | ±3 sec/day | ±1 sec/day |
| Impulse pattern | Single-beat (every other swing) | Double-beat (every swing) | Single-beat (every other swing) |
| Shock tolerance (carry vs bench) | Poor — trips on shock | Excellent — self-aligning | Very poor — must be carried gently |
| Self-starting after stop | No — needs a shake | Yes | No — needs a shake |
| Parts count in escapement assembly | ~6 parts | ~9 parts | ~11 parts |
| Period manufacturing cost (relative) | 1.5× | 2× | 5× |
| Production lifespan in market | ~1780 to 1860 | ~1820 to present | ~1770 to 1950 (marine use) |
| Suitability for wristwatch use | Unusable — too shock-sensitive | Standard choice | Unusable — too shock-sensitive |
Frequently Asked Questions About Duplex Escapement
That's classic duplex tripping under shock. The escapement has almost no margin between the locking-tooth tip and the ruby roller surface — about 0.02 to 0.04 mm. Walking shock is enough to deflect the balance briefly past the unlocking point, the wheel jumps a tooth without delivering proper impulse, and the watch gains a half-turn of the escape wheel for free. Repeat this several times an hour and you get massive rate error in pocket service that vanishes on the bench.
Diagnostic check: lay the watch dial-up on a soft pad and tap the side of the case sharply with a wooden pencil. If you can hear the wheel skip ahead, the escapement is tripping and you need to inspect roller-to-tooth depth. The fix is usually re-pinning the balance to lift amplitude back above 220°, not adjusting the escapement geometry.
Two reasons converged. First, the lever escapement got dramatically cheaper as Swiss factories standardised the club-tooth lever and pin-pallet variants — by 1860 a basic lever movement cost roughly the same as a duplex to manufacture. Second, the wristwatch began emerging, and the duplex is fundamentally unusable on the wrist because constant arm motion trips it within minutes. The lever's self-aligning safety geometry handles wrist shock; the duplex doesn't. Once wristwatches started replacing pocket watches in the 1900s, no rational manufacturer would tool up for a duplex line.
No — and any reputable restoration ethic forbids it. Swapping a verge for a duplex changes the watch's identity, destroys collector value, and usually requires re-cutting the pillar plate to accept a different escape wheel arbor position. If the original verge is beyond repair, the correct restoration is to cut a new verge to the original geometry. The only time a duplex retrofit is appropriate is on a movement that originally carried a duplex but had a later (incorrect) lever conversion done in the Victorian era — there you can argue you're returning the piece to original specification.
A duplex ticks roughly half as often as a lever of the same balance frequency, and the tick is asymmetric — you hear a clear loud impulse beat alternating with a softer dead-swing rustle as the locking tooth slides on the ruby roller. A lever produces an even, symmetric tick-tock at uniform volume. If your duplex sounds symmetric, the impulse pallet depth is wrong and the wheel is delivering partial impulse on both swings instead of full impulse on one — rate stability will be poor and amplitude will sag.
The duplex is a single-beat escapement, so the wheel only advances on every other balance swing. With an 18,000 bph balance (5 swings per second) you get 2.5 escape-wheel-tooth advances per second — and with a 15-tooth wheel geared to a centre-seconds hand through the standard duplex train, that works out to exactly one seconds-hand jump per second. This was a deliberate selling point in the 1820s: buyers associated the clean once-per-second tick with marine chronometer accuracy, even though the duplex isn't actually a chronometer-grade movement.
Target clearance is 0.02 to 0.04 mm measured at the tip of the locking tooth against the cylindrical face of the ruby roller. Open it to 0.06 mm and the wheel will trip on routine handling — the watch becomes unusable in pocket service. Close it below 0.02 mm and the wheel binds when the watch heats up from body temperature, because brass plate expansion shifts the wheel arbor position by about 0.005 mm per 10°C across a typical 30 mm plate span.
If you're struggling to assemble at 0.03 mm, the problem is almost always a bent balance staff pivot or a roller that isn't seated square on the staff — not the clearance spec. Fix the underlying alignment, don't open the clearance.
References & Further Reading
- Wikipedia contributors. Duplex escapement. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
