The tuning fork powered Omegas are amongst the most enigmatic and romantic in the entire Omega history. I recently posted a short and hurried explanation of how these 'hummers' worked for Eric and, as a result he has asked me to write a longer article.
Every timekeeping device needs a timebase. The oldest timekeepers, clepsydra (waterclocks) and gnomon (sundials) are not always suitable for wrist use (as you can see, I have tested this statement. Oh, and the match is still 'live' so you can also use it at night)
There are only three, perhaps four, successful timebases used in watches. The first is the balance and escapement about which I will say no more. The second is the metal tuning fork. The third is the quartz crystal tuning fork and, depending on your taste, the fourth is the lenticular quartz resonator. At some point, the chip scale atomic clock will arrive and I read recently that piezoelectric carbon nanotubes had been experimentally developed. These have my bet as the timebase of the future. I feel sure there are other, unconsidered, possibilities.
The point is that we are poised in a period of utterly unprecedented scientific change in which new technologies can arrive without warning and utterly displace previous technological miracles. Tuning fork watches are a perfect example of this.
The first prototypes were made in 1960 by Bulova and the last in 1977. There was also an incredibly brief resurgence in the nineties which did little more than burn through the spares reserves for the ESA branded, but Bulove licensed 9162 and 9164 spares. No new index wheels were made.
The tuning fork arguably reached its zenith in 1969 when it was selected as the timebase used by the Apollo missions. They may have had Speedmaster on their wrists but they had Bulova Accutrons in their instrument panels and instruments. Indeed, the only 'watch' left on the Moon in '69 was an Accutron which gave a timebase to some instruments left behind. (Not to mention the crystal of a Speedmaster which famously popped off on the Moon).
However, mere months later, on Christmas day 1969, Seiko put the very first quartz on the market and it was all downhill from there.
The first, and most notable thing about a tuning fork watch is the fundamental fact that, inside the watch there is actually a real tuning fork, that is actually producing a stable musical note. (just off the D above middle C in the case of the F300) - as a result, f300 owning mandolin payers need never worry about tuning forks ever again!
Now, a tuning fork works by vibrating at a set frequency, which if tuned correctly, happens to correspond to a musical note. This in turn causes the air around it to vibrate and so on. however, there is nothing stopping a cunning engineer from tuning a tuning fork to vibrate at a convenient frequency for easy maths and engineering...
The challenge with a tuning fork watch is to turn that convenient 300hz hum from a reciprocating to a rotary motion as the kindly Doctor Russell's animated graphic shows:
So, how is this trick accomplished? As usual, a nice picture will help. Using my post credit crunch budget of 17 pence (that's cents in the US or, if you are reading this in three months time, dollars ) I have produced this almost 'computer generated' image:
The tuning fork is a tuning fork, the tweezers are the static body of the watch and the finest, hand whittled, endangered hardwood toothpicks from the tropical island paradise of Edisto, represent sprung steel arms tipped with tiny ruby jewels that engage with the 300 carefully engineered teeth (yep, that's old bluetac) on the index wheel (a coveted HEQ forum badge)
The crucial bit is that the pawl arm is attached to the tines of the tuning fork which moves while the index arm is attached to the body of the watch which is static. As the tuning fork vibrates right, the pawl arm, which is engaged in a tooth of the index wheel, pushes the tooth and thus the index wheel round by slightly more than one three hundredth of a revolution:
As this happens the jewelled end of the static index arm slips back on to the next tooth back. Thus,
when the tuning fork vibrates left the index arm holds the index wheel in position and stops it slipping backwards. Thus it is ready for the tuning fork to begin to move to the right again:
This movement continues back:
In a way that gives a whole new meaning to the words 'watch pornography'. However, the crucial point is that each frenzied thrust of the pawl arm carries the index wheel round a further 300th of a turn while the index arm stops the index wheel from losing control and slipping backwards in the excitement. *Ahem* I really need to get out more.
Obviously, you will also want to see the video...
and for the truly depraved:
you sick, sick chronoperverts! (boy, I hope I don't get banned for this!)
So, that's the basics.
However, what is astonishing is the scale of the parts that accomplish this remarkable little feat of engineering. The smallest tuning fork powered ebauche available, the 23xx has a tuning fork that is just under a centimetre and a half long:
This one is missing the arms, but it gives you an idea of the quality and scale of the engineering going on here! Moving on to the ESA 9162 or, if you prefer, the Omega 1250, you can just see the two arms with the ruby tips.
Can't see them? The toothpick is for scale, but, starting near the tip of the pick move your eye right until you see the ruby jewel. at about four o'clock, just to the right and below the jewel, protruding on to the housing you can see the pawl arm, with the index arm just below it. Remember, the toothpick is for scale and the pawl arm is vibrating three hundred times a second!
Again, the toothpick gives scale. This miniature bridge houses the index wheel. furthest left, yes, it has 300 miniature teeth cut into it.
(rumour has it this is done by highly trained, but underpaid, fleas)
Yes, it is about the same diameter as the toothpick. That is why you cannot get them made up any more. (the fleas died)
Here are the two wheels in position. As I learned the hard way, you have to fully replace the bridge and then gently reset the indax and pawl arm into place from the other side, through a hole about the size of two toothpicks.
If you try to do so before the bridge is in place, well the arms are made of very very high quality spring steel and the index wheel is very very light. (and you cannot use a magent to find it as that magnetises it and that cause some very odd behaviour once in the watch again.) These mechanisms are excitingly easy to break!
In fact, so as to avoid too much disappointment, let's have a brief reality check here. While a working movement is remarkably robust, some of the engineering here is quite astonishingly fragile when confronted with, well, not to put to fine a point on it: me, you and any watchmaker untrained on these increasingly rare movements...
First the obvious stuff.
When changing the time there is a little tiny slip clutch that stops you spinning index wheel around. Obviously this is a Bad Thing. This clutch is, itself, not that robust as it relies on being well lubricated. when it stops slipping then every time you adjust your watch you are spinning the index wheel. The practical effect of this is that you are rapidly stripping the microscopic teeth off of the index wheel. Like this:
Obviously, no teeth means no movement. It will hum even more smoothly than before but it's dead and they don't make index wheels any more.
Next, even the finest sprung steel suffers metal fatigue, and those little jewels are only stuck on with epoxy. Once these have snapped or dropped off (or been ripped off) the watch is, again,very harmonic scrap, like so:
In addition to this they don't make any parts of the circuit any more so the coil is another area to fear. In short, when you buy a non runner or one that is humming strongly but without any movement, then remember that you are most probably buying the equivalent of this:
rather than this:
and bid accordingly!
So having brought you all down to earth with a bump, allow me to point out some of the wonderful aspects of this magnificent movement. The best thing about the f300 is that the process described above provides both the timebase and the motive power for the rest of the movement, which gears down the 300 tiny steps to a seemingly absolutely smooth sweep. This allows these ebauches to be very very simple and incredibly robust (in normal, correctly serviced use)
The human eye is capable of detecting down to about 100ms, or ten steps a second, as found in the old ultra high beat 36000 movements. However, whenever the eye detects two flashes, (and to the eye, the resting point between ticks is, functionally, a flash because the actual movement is too fast and small to register.) it will perceive those two flashes as movement between the earlier and later flash. (to be precise, any number of flashes up to 2 degrees of visual angle and 200ms or less apart will be seen as movement - weird eh?)
This is called the phi phenomenon and is the underlying cognitive mechanism behind everything from cinema to disco light ropes. It also gives neurobiologists a miserable time, cognitive scientists research grants and philosophers something impressive to bullshit about at parties.
It is this optical illusion that allows Rolex to claim that their movements have a sweeping movement. however, they don't. The f300 on the other hand does. While it is in fact ticking at 300 times a second this is way, way, way below the eye's ability to discriminate and is genuinely perceived as smooth movement. Only two watches have a smoother movement: the Seiko Springdrive and the legendary Omega F720 Megasonic.
Now, Seiko would have you believe that the Springdrive is utterly smooth, but this isn't true. While it is true that the Springdrive's hand never actually physically stops, it isn't very smooth. This is due to the digital control circuitry, which samples and breaks the glide wheel only a few times a second. the upshot of this is that the second hand of the SD is constantly accelerating and decelerating. Once again, the eye tricks us as it is brilliant at spotting movement but rubbish at spotting acceleration and deceleration, especially at that scale.
The Megasonic, on the other hand, 'ticks' 720 times a second. First, this is ridiculously smooth in its own right. Secondly the first few gears are not mechanical but magnetic. As a result, the pulses are smoothed asymptotically and, to all intents and purposes, cease to exist. Thus, the award for smoothest second hand of all time, goes to the Omega Megasonic:
Oh? you want me to open it so you can see the back too? I spoil you!
See that little box with a wheel in it at about 1:30? that's the 'souris' or 'mouse' and is the only other technique for converting vibration into rotation. it is completely sealed, and as this earlier picture shows:
it's about the size of a section of toothpick. It is also mounted directly upon the tuning fork itself, hence the magnetic gearing! God knows how they made it but they don't make it any more!
When it's running, it looks like this:
you also have to turn the sound up to get an idea of the unholy wail that this, clearly haunted, movement emits. I am genuinely not exagerating when I say that you will distress dogs and cats at fifty yards! I can also guarantee that you will only sleep with it on once. Even then, it will not be for long.
In the case of this movement there are two arms which simultaneously perform the act of pawl and register arm allowing the movement to work while having no fixed index point. Here's a video that demonstrates how this works so well that I really don't feel the need to explain further.
Right, that's enough for me. I'm off for some tea.
However, having spent about four hours on this any comments would be appreciated...