The fascinating tale of clepsydra, pendulums and quartz
Clepsydra, pendulums and quartz? Have I some lock-down cabin fever? A sake too far?
Well with one thing and another, I have been favouring my simple quartz watches lately. In this I suspect I am not alone.
And this is mainly a long ramble. At least a two-cuppa. I can’t guarantee factual accuracy. It’s not a thesis. There are doubtless many mistakes. And I’ve probably already blown my avoiding-contention budget in choosing pendulums over pendula. With all that understood, prepare your horological digestive for a long dunk into the hot mug of quartz history.
My very first watch was a quartz, some time in the 1970s, and a new-fangled digital LCD type at that. These were innovative and accurate and very modern. They were all set to brush away our dads' old-fashioned and inaccurate mechanical watches into well-deserved and permanent obscurity. James Bond had moved on to digital from his watches-of-yesterday too, those creaky old Rolex Submariners and 6238s.
Pah, who'd be seen with one of those, when you could have one of these:
https://live.staticflickr.com/65535/...7fca9f87_c.jpg
Yes there is a watch there
In the absence of other influencers, this truly vindicated my young self’s excellent choice to move with, well, the times.
There was no such thing as a “child’s” digital watch, so my treasured item, by Ingersoll if I recall, hung proudly off my arm like a manacle.
https://live.staticflickr.com/65535/...fb93951f_c.jpg
Not my actual watch. Picture handcuffs on an action man
I was though, briefly, the toast of my primary school. Until Trevor or Derek or some other 70’s named boy bought one that featured the seconds readout on the same display. And then Alan or Brian or Richard acquired one with a stop watch with a 100th seconds counter, and perhaps even a little marker denoting the day of the week. The game truly was afoot. It was a time when I could feel technology accelerate away from me.
My fascination has never subsided, although it took a deep detour into mechanical watches for a long time.
But innovation in the mechanical realm, at least during my lifetime, has tended to be a new bezel colour (“stunning”) or minor variation in case dimension or a re-issue of an older watch design. All good for Instagram appreciation and forum rumination. Or for rewarding one’s attainments, taste or discernment. Perhaps in some cases, even for telling the time.
But notwithstanding a newer escapement here nor a longer mainspring there, I think that most progress, viewed objectively in horological rather than jewellery terms, has been in the quartz realm for most of my life.
And somewhat before my life too...
(You are going to need that biscuit, perhaps even a packet. You deserve it)
https://live.staticflickr.com/65535/...cf71a68b_c.jpg
As generally understood, Seiko’s Astron was the first commercial quartz watch, released on Christmas Day in 1969.
https://live.staticflickr.com/65535/...25c7cc4215.jpg
Seiko 35A movement from the Astron. Genesis. It took a bit more than 7 days though
But quartz time pieces (rather than wristwatches) had been around since the late 1920s.
Canadian Warren Marrison and his Bell Labs colleague J.W. Horton are credited with inventing the quartz clock in 1927:
https://live.staticflickr.com/65535/...34643e22_z.jpg
Hodinkee "Marrison" tribute version expected imminently
The keen-eyed will have already spotted the obvious challenge: how to fit it onto a Nato strap?
So just how did the quartz clock become the quartz watch? We have to go back in well, time, a little.
https://live.staticflickr.com/65535/...cdf87646_z.jpg
19th century illustration of 3rd century BC clepsydra
As with this water clock, there is something of a continuous flow through horological history from those ancient Babylonian clepsydra to the GPS quartz watches today. Sometimes development runs quickly, other times it meanders, awaiting a new direction or the gradual erosion of an obstacle.
Water clocks had been around for at least a couple of thousand years before they were finally eclipsed technologically by mechanical clocks in the 10th century.
These in turn were much improved by the invention of the escapement in the 14th century.
A clock built in this time by Henry De Vick for Charles V is still in use (with quite a few subsequent services and upgrades, naturally…) in the Palais de Justice in Paris. Imagine!
https://live.staticflickr.com/65535/...9864893c82.jpg
Best not enquire about service history
The escapement was the first time that vibratory motion in a mechanism was used to control the rate of a clock.
It’s a fair bet that whatever watch you are wearing today, quartz or mechanical, depends on vibration rather than falling weights or the flow of water, to measure the passage of time.
The next improvement was in regulating that escapement using a resonant element. Resonance being either mechanical or electrical, whereby deformation from rest results in a returns to the initial position. Or as physicists might describe it, potential energy changed into kinetic energy, with a little lost in each oscillation due to friction.
The pendulum fits that description. Its use as a resonant element to regulate a clock was described by Galileo Galilei to his son Vincenzio sometime around 1637:
https://live.staticflickr.com/65535/...0de7050a_c.jpg
The period of oscillation, Galileo observed, was seemingly constant regardless of amplitude. An excellent attribute for timekeeping.
This concept of using resonance for regulation remains in use in mechanical (balance resonator), quartz (crystal resonator) and atomic (atomic resonator) time pieces today.
Galileo was near blind at the time, and his idea was not built. But a working model was later fashioned from these drawings, and can be seen in the Science Museum in London.
The first actual use of a pendulum in a clock is attributed to Dutchman Christian Huygens, in 1657:
https://live.staticflickr.com/65535/...afa820e1_z.jpg
Not just tulip aficionados
Moving on, but still with resonance, French physicist Jules Antoine Lissajous showed in 1857 that a tuning fork could be sustained in vibration indefinitely by electrical means, using an electromagnet. We are getting closer!
Although the principle of maintaining resonance at a fixed frequency is common to mechanical and quartz watches, using electricity requires less energy, is more consistent, and has negligible wear, compared to a pendulum (or balance wheel) mechanically stimulated.
And so, finally, to quartz as a resonator.
In addition to its physical and chemical stability, the elastic hysteresis of quartz is extremely small. It requires only a tiny amount of energy to sustain oscillation.
A good tuning fork (in a vacuum) might resonate around 2000 times before reaching half its original amplitude. The best pendulums might manage up to 20,000 times. But a quartz crystal can achieve over 1,000,000 vibrations before falling to half amplitude. Due to the piezo-electric property of quartz, its frequency can also be determined electrically too.
Thus following Marrison and Horton’s quartz clock of 1927, all the processes necessary for the quartz watch existed - an accurate resonator and a means of maintaining and measuring the vibration (electrical).
Quartz clocks of these times were generally only used in labs. They were driven by vacuum tubes and power-hungry frequency dividers reduced the resonant frequency to something more easily coupled to a synchronous motor. Later, as transistors replaced tubes, table-top “electric” quartz clocks became viable.
The main challenges in making a quartz watch were to reduce its overall footprint and to lower its power consumption. The total size needed to be around 3cm cubed or less, and power consumption less than 10μW. This was to enable one year of operational life from a 1.35V mercury button cell compatible with the watch volume.
In 1962, it was shown by Eric Vittoz at CEH (Centre Electronique Horloger, a lab set up that year by the Swiss watch industry in Neuchâtel) that a commercial 10Khz quartz crystal could be used, with four transistorised divide-by-ten frequency dividers, to output a signal at… 1Hz.
https://live.staticflickr.com/65535/...998c5a8b_c.jpg
A demonstration of Vittoz's transistorised divider
Yes, 10KHz divided by 10 four times gives us the precious one-cycle-per-second. An electronic tick!
And furthermore, the whole circuit consumed less than 10μA at 1.5V. It was true that the actual display took more power still. But it was close enough that, as Vittoz himself recalls, a 1.8 million Swiss Franc research budget for 1963 was approved by the CEH board of directors.
https://live.staticflickr.com/65535/...5d44323f_c.jpg
From there it was a race to further reduce power consumption to meet the 10μW goal for the entire solution, display included. Engineers from around the world were engaged.
There were many dead ends. One of the design requirements was for a quartz frequency of 10Kz or less, simply to ensure the fewest number of power-consuming divider circuits. Another was for that crystal to have a precision better than 10 parts per million (roughly 1 second per day).
But commercial quartz crystals of these frequencies and quality were simply too large. Eventually, Armin Frei, also working at CEH, managed to mount a miniaturized 8KHz quartz in the vacuum of a small metallic package:
https://live.staticflickr.com/65535/...736a7da3_c.jpg
This is the actual prototype
Which led to focus on the frequency divider. Again, many approaches were tried. Diodes. Magnetic flux accumulation (similar to computer memory circuits of the time) and capacitors. These were analogue circuits and the maximum frequency division was limited by the precision of the components, generally a “divide by 10” being about the most consistently usable.
Vittoz subsequently created a newly-developed frequency divider of his own design based on a phase-lock-loop circuit:
https://live.staticflickr.com/65535/...9921b729_c.jpg
Still not quite ready to wear
It probably says something about the challenge, that when presenting such ideas at the 1964 International Solid-State Circuits Conference, the concept of miniaturising all this into something that could fit into a watch was considered foolish.
"Foolish" is engineer speak for “near-impossible, best start looking elsewhere for next years research budget”.
Persistence and two years later however, a frequency divider based on binary division (i.e., divide by two) was developed. There were no circuit simulators nor chip design tools, so the proposed design was “integrated” by first making a larger version (multiplying component values by 1000 for example) on a breadboard for modelling, before the smaller version was burned into silicon.
Here’s the first such divider integrated circuit used in a watch:
https://live.staticflickr.com/65535/...1d1f056076.jpg
1966: getting closer to size and power requirements now
This contains 110 components and yet permits the power draw, including the frequency dividers, to be 12μA at 1.3V.
The first prototype (Beta 1) had used that 8Kz crystal, with 13 binary dividers (8KHz->4KHz->2KHz->1KHz->512Hz->256Hz->128Hz->64Hz->32Hz->16Hz->8Hz->4Hz->2Hz>1Hz) to drive a synchronous motor once per second.
The second prototype, Beta 2, used just 5 binary dividers resulting in a 256Hz output frequency. This was to reduce the power needs of the circuit. The output drove a motor and through a gear train, the frequency was reduced down to show the time. Ten of these prototypes were created in 1967.
https://live.staticflickr.com/65535/...fb2bcfb5_c.jpg
Size and power objectives met!
The commercial version (Beta 21) was modified further, with such additions as a rate trimmer capacitor. About 6000 examples were made to be sold under various Swiss brand names including Omega, Rolex and Patek Philippe. The Beta 21 was launched in 1970, just shortly after Seiko’s Astron.
Today most quartz watches operate using a 32KHz crystal (there are notable exceptions) with a 20-stage divider. A good quality 32Khz quartz crystal can still be physically small enough (about 3mm) while lower power divider circuits permit a higher base frequency.
The advantage is the elimination of the rate trimmer, which had proved problematic in the Beta 21. Fine-tuning is obtained by removing pulses (“inhibition”) before they enter the divider. With a 20 stage divider, the average frequency could be adjusted to within 1 part per million, or roughly 0.1 second per day, depending on the performance of the actual circuit.
And some examples:
https://live.staticflickr.com/65535/...966831fd_c.jpg
Omega Beta 21 (not mine)
https://live.staticflickr.com/65535/...db7f677e_z.jpg
Omega Megaquartz (the inside of mine... for the outside, try here
https://live.staticflickr.com/65535/...e7f45cb9_z.jpg
My modern Astron (yes it fits on a Nato)
All quartz watches share this rich and exciting history. Many interesting ones don't cost much either. For examples, see the interesting quartz thread.
If you made it this far, congratulations! I hope you enjoyed this little meander through a truly fascinating history. A time of horological innovation, and one we are actually living through. I find wearing even my cheapest quartz watch provides a source of great pride in the ingenuity of others. I appreciate the no-nonsense classless appeal, and the absence of associated stresses.
I think in these times especially, there is some accessible comfort in this.
Paul
