watch elements 20180504

Version 20180504
Essential pieces of a watch
Traditional watches all include the following basic elements:
1. A timekeeping element
Some element that keeps track of time.  In a traditional watch, this is some oscillator that maintains a constant frequency.  That constant frequency ultimately is what is used to keep track of time.  Common oscillators include hairsprings, tuning forks, quartz crystals, and cesium atoms, altohugh other mechanical oscillators may be used.
Oscillators are typically measured in Hz ("cycles" per second).  Note that the frequency in Hz of an oscillator typically refers to how long it takes for the oscillator to return to an initial state.  In many watch designs, the dispay will advance twice over the course of one cycle.  For example, in a traditional mechanical watch with hairspring and lever escapement, the hands will advance twice for each back-and-forth of the balance wheel (this is why hairspring/lever watches are often quoted in "vph" - vibrations per hour - rather than Hz.  Such a watch that runs at 4Hz makes 2*4*3600=28,800 vph).  All of the oscillators we consider in this analysis will be quoted in Hz.
A traditional hairspring oscillator most commonly runs between 2Hz and 4Hz.  Some less common hairspring oscillators run as low as 1Hz and as high as 100Hz.  The highest-frequency hairspring I'm aware of in a production watch that is used for timekeeping (as opposed to a specialized chronograph-only second gear train) runs at 10Hz (Breguet).
Very recently, LVMH (Zenith, Tag Heuer, etc) has introduced an oscillator in the form of a solid piece of deformable silicon.  The silicon is bent slightly out of shape when a force is applied, and then returns to its original shape in a precisely-tuned amount of time.
Tuning fork oscillators were developed in the 1960s.  These run at considerably higher frequencies than hairspring oscillators and keep more accurate time as a result.  These oscillators run as low as 300Hz and as high as 720Hz.
Quartz oscillators were developed for wristwatches in the late 1960s and early 1970s.  These run at still higher frequencies than hairspring oscillators.  The first quartz wristwatch oscillators ran at 8KHz.  Common modern quartz oscillators run at 32KHz.  The highest frequeny production quartz oscillator runs at 262KHz (Bulova).  Very high accuracy quartz watches were previously developed running at frequencies up to 4MHz.  Most modern high-accuracy quartz watches use thermocompensation to achieve high accuracy, rather than ultra-high frequency, altough Citizen has recently announced (but not released) an 8.4MHz quartz watch, rated at 1spy.
Note that all of the oscillators presented above operate, on some level, in an essentially similar way - some physical object (spring, piece of silicon, tuning fork, quartz crystal) is stretched out of its natural resting shape, and then returns to its natural resting shape in a carefully-controlled amount of time.  Very recently, chip-scale atomic clock (CSAC) oscillators have been available for wristwatches - these are the only wristwatch oscillators that operate on a different principle.  These use the same technology used in atomic clocks and are accurate to 1 second per millenium.  They feature a miniature oven in which cesium is allowed to transition between quantum energy states; when this happens the atoms emit light with a known, fixed frequency.  The oscillator (which is actually the changes in the EM field caused by the emitted light) "oscillates" in the gigahertz range.  These watches are still quite uncommon, quite large, and quite expensive - as of May 2018, none are in current production, with Hoptroff recently ending production of watches.
As an aside, regarding oscillators - while all wristwatches that I am familiar with use oscillators, not all timekeeping devices do.  Sundials, candle clocks, and clepsydra (water clocks) measure time continuously.  Some oscillators that are present in clocks that are not present in watches include pendulums and torsion pendulums (JLC Atmos, for example).
2. Escapement
Here we use the term "escapement" loosely.  Traditionally, an escapement refers to a mechanism in a mechanical watch that prevents the gear train from advancing except for fixed amounts at fixed intervals (governed by the hairspring oscillator).  Here we will use it to describe whatever mechanism translates the timekeeping of the oscillator to the rest of the watch.
Not all escapements translate every "beat" of the oscillator into an advancement by the rest of the watch.
The most common escapement in traditional mechanical watches is a Swiss lever espacement.  This involves a rotating balance wheel, a fork, and an escape wheel.  The teeth of the fork block the escape wheel from turning.  As the balance wheel spins back and forth, a jewel attached to the wheel pushes the fork left and right; each time allowing the escape wheel to advance.  This escapement advances the hands twice per complete cycle of the oscillator (in other words, a 4Hz hairspring with a Swiss lever escapement would see the hands advance eight times per second).
There are a number of similar escapements to the Swiss lever escapement.  The Daniels co-axial escapement, GP Constant Force escapement, detent (or "chronometer") escapement, and others all work in essentially the same way.  One significant difference can be the advancement frequency; some, like the Daniels co-axial escapement, advance the hands twice per cycle (like the lever escapement); some, like the detent escapement, advance the hands only once per cycle.
These escapements are overwhelmingly more common with hairspring oscillators.  There are a few exceptions which pair this style of escapement with a quartz oscillator - the Luch 3055, a Belarusian watch from the Soviet era, which uses a balance-wheel based escapement (I do not know which) along with an IC attached to a quartz oscillator, the Timex model 62 and 63, which work similarly, and some others which are extremely obscure.
By contrast, most quartz-oscillator analog watches use an integrated circuit and stepper motor as the "escapement".  The integrated circuit counts the cycles of the quartz crystal, and at appropriate intervals, activates an electromagnetic stepper motor, which advances the gear train.  While obviously radically different from the balance-wheel escapements described above, this part of the watch peforms largely the same function, translating the "back-and-forth" behavior of an oscillator into uniformly-advancing motion.  While the balance-wheel oscillators may advance the hands once or twice per cycle, an IC/stepper motor attached to a quartz oscillator will typically advance the hands once per second - or once per 32 thousand cycles for the most common quartz oscillator frequency.  There are some stepper motor watches that advance the hands more than once per second (Seiko 5S21/5S42, Sieko mecaquartz, Bulova precisionist), but the cost of more frequent stepper motor activation is faster energy consumption.  A thermocompensated IC/stepper motor escapement with quartz oscillator also includes a small thermometer, which is used to vary the number of cycles counted before advancing the hands, in order to account for the fact that the frequency of a quartz oscillator varies with temperature.  This is also the only escapement design currently in use for CSAC oscillators.
LED/LCD digital watches operate much the same way, but instead of activating a stepper motor, the IC simply changes what is displayed.
Yet another kind of escapement, found only with quartz oscillators, is the Spring Drive.  To my knowledge this is found only in Seiko Spring Drive movements and in the Piaget 700P movement, but the Seiko is so much more common, more accessible, and most aggressively marketed that the escapement design is widely referred to simply as the Spring Drive.  Here, the escapement consists of an IC that counts the cycles of the quartz oscillator, and an electromagnetic brake that maintains the speed of a wheel's advancement.  It shares much more in common with the design of a traditional Swiss lever escapement than a stepper motor-based escapement does.  Instead of a wheel (gear) being physically blocked and unblocked by a fork, it is slowed by an electromagnet.  In principle, the same escapement design could be achieved with an oscillator other than quartz (for example, an IC that used a camera to count the oscillations of a hairspring and then use that information to electromagnetically maintain the speed of a glide wheel), but this has not been done.
The final escapement design we will discuss is the index wheel.  In this design, most common with tuning fork oscillators, one "finger" (typically a jewel on a pawl" holds a sawtoothed-wheel loosely in place and prevents it from turning.  A second finger is mechanically attached to an oscillator (i.e., tuning fork).  As the moving finger moves back and forth, it pushes on one of the teeth of the wheel, causing it to advance.  The stationary finger slides up one of the teeth and then slips over the top, preventing the wheel from moving backwards.  The moving finger then moves back over the tooth it pushed on, and the prcess begins again.  All tuning fork-based watches use this type of escapement.  Many early quartz watches also used an index wheel escapement (for example, the Beta 21 and the Longines 6512).
As an aside, the index wheel escapements that were produced for tuning fork watches in the 1960s and 1970s are some of the most incredible pieces of microengineering of the era.  The tolerances of the teeth in the wheel are unbelievably fine, and to this day the process for creating the wheels is a trade secret.  No new wheels are being produced and often the only source of replacement is a donor movement.
3. Energy
All watches require energy to run.  The energy is used to power the oscillator, the escapement, or both.  All watches store energy either in a mainspring or in the form of electricity in a bettery or capacitor.
An aside for the purpose of clocks: many clocks store energy not use a mainspring or electricity, but using the gravitational field of the earth.  The clocks are powered by weights falling, and recharged by people winching the weights back up.  In terms of energy consumption, these clocks typically work in a way not too different from mainspring-based systems.
The most obvious way for a watch's energy to be recharged is for the person wearing it to do something.  These manual-wind watches typically allow the wearer to recharge the energy by turning the crown of the watch, although some ma be rewound using a key (Lange 31) or a button (Romain Gauthier Logical One).  I am not aware of any watches with a manual recharge where the energy is stored as electricity.
Watches can also be recharged through motion.  A rotor moves as the watch's orientation in the gravitational field of the earth changes, recharing the watch.  Mainspring-based watches in the 1950s used spring bumpers that translated this motion into a rewinding of the mainspring; later more sophisticated system used a ratchet more directly connected to the rotor.  This kind of energy recharge is also found in watches that store energy as electricity - for example, Seiko Kinetic watches.
Electricity-based watches can be recharged by solar power.  Citizen is a major driver of this technology.
Electricity-based watches can also be recharged using an external power source (i.e., plugging it in).  All current CSAC watches use this approach.
Finally, the energy in a watch may be recharged by simply removing a depleted energy store and inserting a charged one.  This is commonly done with batteries in electricity-based watches; as far as I am aware, there are no watches designed fo the frequent removal and replacement of mainsprings.