Modern wristwatches are often thought of as luxury accessories, in a world where GPS-corrected electronic timekeeping is ubiquitous. Buyers of today might value a watch for its beauty or style, or perceptions of exclusivity and status, but rarely for its ability to tell the time accurately. And that’s a real shame, because the history of watchmaking is all about the never-ending pursuit of precision, and by ignoring that heritage we lose a large part of what makes watches special.
The advancement of timekeeping technology was once of paramount importance, particularly in the endeavours of long-distance travel. History is rife with examples of the deadly consequences of poor timekeeping. The Scilly naval disaster of 1707 sent 2,000 souls to a watery grave, and was attributed to the ships navigator’s inability to calculate their longitude without accurate clocks onboard. This led to the invention of the Marine Chronometer by English clockmaker John Harrison, a clock of sufficiently stable rate to serve as a portable time-standard at sea, which greatly increased the safety of sea travel.
Another example occurred in 1891 near Kipton station in Ohio, when a train conductor’s pocket watch stopped for four minutes, leading to a full-speed collision of two trains that killed nine people. This led to the establishment of Railroad Pocket Watch standards by Webb C. Ball, ensuring that conductors had sufficiently reliable timekeeping instruments and creating the modern catchphrase of being “on the ball”. These may be events from the distant past, but they serve to highlight something we seldom appreciate about horology in the modern era: the ability to know the time accurately was literally a matter of life or death for our forebears.
These examples also demonstrate that watches had to be isolated sources of truth for most of their history. While we take for granted our constant connections to the Internet and satellites equipped with atomic clocks, as an IT professional by trade I know how fallible such reliance can be. Smart watches have their appeal, but a big part of why I’ve become so attracted to more traditional watches is that I don’t want to worry about whether or not my watch needs a software update, or has security vulnerabilities that need to be patched, or is failing to sync with GPS, or isn’t pairing via Bluetooth, or is lacking 4G connectivity, or is collecting and selling my personal data. A watch without these uncertainties or reliance on external signals to tell the time accurately is one less thing to stress over, and the ultimate expression of hassle-free autonomous precision is high accuracy quartz, or HAQ for short.
To fully appreciate the engineering marvels of HAQ, we first have to understand how quartz watches work and how they have improved over their predecessors. The oscillator within a quartz watch is a small quartz crystal, usually shaped like a tuning fork (the technical term for this is an XY-cut crystal). Quartz is a piezolelectric material, which in layman’s terms means that if you apply an electric voltage to quartz, it deforms (piezo means to squeeze in Greek). This squeezing and relaxing of the tines of the crystal tuning fork make it vibrate at a very constant rate, and a circuit counts these vibrations and advances the second hand via an electric stepper motor each time the count reaches 32,768, the standard frequency used for quartz watches.
The ability of a quartz crystal to vibrate so fast – 32,768 times per second – is one of the main advantages it has over the traditional hairspring. A modern mechanical hairspring typically operates at 4Hz, aka 4 oscillations per second, so the crystal in a quartz watch is moving over 8,000x faster than the hairspring and balance in a mechanical watch. This benefits timekeeping in two ways. Firstly, it allows for much greater precision in calibrating the oscillator, being able to adjust the time in thousandth-of-a-second increments rather than quarters-of-a-second. Secondly, it makes the oscillator much less prone to interference via shocks, the wearer’s wrist movements, or loud noises, as 32KHz is a frequency much farther removed from everyday vibrations than 4Hz. You know how you can feel the bass make your chest vibrate in a live concert? Low frequency sounds like that can interfere with something vibrating at a similar frequency (like a mechanical hairspring and balance), but wouldn’t make a jot of difference to a much higher frequencies of a quartz crystal.
Mechanical oscillators also suffer from a number of potential error sources – positions, isochronism, magnetism, temperature, and aging – for which quartz is (mostly) unaffected. Positional error, aka the drag of gravity causing the hairspring to run faster or slower depending on its orientation towards the Earth, cannot affect the stiff tines of a quartz tuning fork to anywhere near the degree it affects the thin coils of a hairspring. Isochronism or the tendency for a hairspring to oscillate slower as the mainspring unwinds and torque is reduced isn’t a problem for a quartz crystal with a regulated power supply. Quartz is also a non-ferrous material, and thus is completely unaffected by magnetism, although modern mechanical watches have increasingly started adopting non-ferrous materials as well like silicon, Rolex’s Parachrom and the Swatch Group’s Nivachron to combat the increasingly prevalent effects of magnetism.
That leaves temperature and aging, which are challenges for both quartz and mechanical oscillators. As ambient temperature increases, oscillators tend to vibrate faster, and as things get cooler they slow down. Mechanical watches attempt to combat this by using temperature-insensitive alloys for their hairsprings, silicon being a prime example once again, but quartz watches can’t borrow the same trick; in order for their oscillator to work, it has to be made of quartz for the piezoelectric effect to trigger. Instead, quartz watches combat the effects of temperature a different way, via thermocompensation (more on that below).
One final potential error source for both watch technologies is aging. In a mechanical watch, this is largely due to the increased viscosity of lubricating oils over time resulting in greater friction and eventual damage to the mechanism. In a quartz watch, aging can be caused by a number of factors, such as contaminants sticking to the crystal and increasing its mass, or contaminants diffusing away from the crystal and reducing its mass, or the electronics in the circuit around the crystal degrading over time.
To summarise, quartz watches have one big advantage over mechanical watches (greatly increased frequency), as well as several eliminated weaknesses (positional error, isochronism, magnetism) and a few shared weaknesses (temperature and aging). In order to push the boundaries of what quartz is capable of, HAQ watches usually either seek to capitalise on the main strength of quartz technology by increasing frequency even further, or they seek to eliminate one or both of the remaining weaknesses in the form of active temperature compensation, and pre-aging their crystals.
Two early examples of the ultra-high-frequency method are the Omega Marine Chronometer from 1974, and the Citizen Crystron 4 Mega 8650A from 1975. Unlike the standard XY-cut tuning fork shaped crystals, both of these watches utilised AT-cut quartz crystals, which are solid blocks of quartz capable of much faster oscillation speeds. The Marine Chronometer operated at 2.4Mhz, while the Crystron almost doubled that at 4.19Mhz, or 4,194,304 vibrations per second, making it 128x faster than a standard quartz watch and over a million times faster than a mechanical hairspring and balance. This allowed the Omega Marine Chronometer to achieve a rated accuracy of 12 seconds per year, while the Citizen Crystron was rated to an even tighter tolerance of 3 seconds per year, making it the most accurate watch ever made at the time. Both of these watches suffered from poor battery life (a year or less) due to the power demands of such high frequencies.
Meanwhile, early attempts at combating temperature’s effects on quartz crystals resulted in watches like the Rolex Oysterquartz, which used a method referred to as “forced constant frequency”. This involved using a compensation circuit that varied the voltage to the crystal as temperature changed, attempting to keep the crystal’s frequency linear. This approach had some drawbacks; most notably, it was very difficult to precisely pair the compensation circuit to the crystal, with small mismatches leading to large variations in timekeeping accuracy.
A more modern method for temperature compensation is referred to as the “digital count adjustment” method, which has become the de-facto standard for HAQ movements today. In this method, the crystal’s frequency is allowed to drift with changes in temperature, but a thermistor measures the change at intervals and references a look-up table within the integrated circuit that knows exactly how much the frequency of that crystal will drift at each temperature. It then adjusts the oscillation count accordingly to advance the second hand early (if the crystal is running slower) or late (if the crystal is running faster). This is simpler in execution than the forced constant frequency method, but it also comes with a limitation: the more frequently the thermistor checks the temperature, the more power it drains from the battery.
A great example of how big an impact this makes is by looking at the differences between the Grand Seiko 9F and the Citizen A060 movements, two high-end modern HAQ movements that both utilise the digital count adjustment method. The Grand Seiko 9F attempts to eliminate all of the traditional weaknesses of quartz by pre-aging their crystals, and then sealing them in a vacuum-sealed cabin within the watch to prevent contaminants from changing the crystal’s mass over time. The 9F also uses active temperature compensation with a thermistor checking the movement 540 times per day, or once every 2.6 minutes. With these measures, the Grand Seiko 9F achieves a rated accuracy of 10 seconds per year, better than the Omega Marine Chronometer with its AT-cut crystal and ultra-high frequency.
Citizen’s A060 doesn’t pre-age its quartz crystal or vacuum seal it, but it has one major advantage over the Grand Seiko 9F: Eco-drive solar power. Because the A060 can continually be recharged by light, it has more power to play with, allowing the thermistor to check temperature more frequently without having to worry about draining the battery. The Citizen checks for temperature 1440 times per day, or once every minute, and through this improvement alone it is able to achieve a rated accuracy of 5 seconds per year, twice as precise as the Grand Seiko 9F.
Then finally, we have the culmination of the HAQ pursuit of precision: the Citizen calibre 0100. This movement uses an AT-cut crystal just like the Omega Marine Chronometer and Crystron 8650A from the 70s, but with an even more insane frequency of 8.4Mhz (8,388,608 vibrations per second). Eco-drive solar power makes this possible, combating the power drain of such a high-speed oscillator by allowing continual recharging of the solar cell. The calibre 0100 pairs this with active temperature compensation (at Citizen’s aggressive once-per-minute intervals), as well as a pre-aged quartz crystal that has been tested and measured during the aging process to ensure the look-up table in the integrated circuit within the watch knows exactly how the crystal’s frequency will drift over time. As a result, the calibre 0100 achieves a rated accuracy of 1 second per year, besting their own record set by the Crystron in 1975 as the most accurate quartz watch in the world.
Article first posted on Hailwood Peters