H.Moser & Cie - Perptual Double Hairspring
In a watch, the gravitational error disrupts the stability of its accuracy. This means that when the watch is in a horizontal position - i.e. when the dial is at the top - the watch runs differently than when the dial is positioned vertically, i.e. laterally. The watchmaker counteracts this phenomenon by adjusting the watch so that the slower vibrations of the balance in vertical position are compensated for as precisely as possible by the more rapid vibrations in the horizontal position. This produces an average level of accuracy, although this is dependent on the way in which the owner wears his watch. This is because the average level of accuracy is only achieved if both positions occur with approximately the same frequency. In practice, the vast majority of watches are therefore adjusted in fi ve standardized positions in order to minimize the discrepancies. At Moser we pursue this practice to the limit, as we adjust our watches in all six possible positions.
So where, then, does this gravitational error come from? And how can we reduce its impact even further? These questions have preoccupied watchmakers for many decades. In this respect, it is important to know the type of attachment at the outer end of the balance spring. This is realized either by a fl at curve or a Breguet curve.
The socalled fl at curve, which is used in most mechanical watches, can be implemented very easily. It does, however, present the disadvantage that the balance spring assumes an asymmetrical shape when it vibrates. In doing so, the centre of gravity in the spring moves away from the middle. If we now imagine that the spring adopts a vertical position, it then becomes clear that due to the shift in the centre of gravity, the vibrations in the "downward" direction are accelerated by the earth's gravitational pull. At the same time, vibrations in the "upward" direction are impeded, and consequently slowed down. On the other hand, if the spring is in a horizontal position, this effect does not play a role. This is not a good starting position for the stability of the accuracy.
In contrast, the Breguet curve was developed in order to avoid precisely this asymmetrical vibration of the fl at curve. It does this by ensuring that the outer end is curved upwards over the high edge, and is then further curved inwards. This task demands all the skills and ability of the adjusters, since the procedure is predominantly carried out by hand. This complicated and from the point of view of craftsmanship highly demanding form of manufacture is the reason why the Breguet curve is only found in watches of very high quality. As a result, the vibrations of a spring with a Breguet curve are almost completely symmetrical. But only almost. There is still a small residual error.
The approach to the double hairspring escapement with a pair of Straumann hairsprings is now very simple: the springs are arranged one above the other, with one vibrating to the left and the other vibrating to the right. If both springs have the same mechanical characteristics, then when they vibrate, the centre of gravity moves outwards from the centre - just as it does with a single spring and a fl at curve. However, as both springs are vibrating asymmetrically in opposite directions as a result of the different direction of rotation, the centre of gravity, on average, remains exactly in the centre. A gravitational error due to the asymmetrical vibrations of the springs can therefore not occur in the fi rst place.
So why was a tourbillon invented to compensate for the error caused by gravity?
At first, the tourbillon was developed to compensate for the gravitational error in a cut bimetallic balance with a steel spring. With this type of balance, the effect of temperature on the accuracy of the watch was prevented by the individual arms of the bimetallic balance bending outwards or inwards with changes in temperature. It is easy to see how the two arms of the balance never moved in a uniform fashion, which invariably caused a much greater gravitational error than the asymmetrical vibrations of a spring with a fl at curve. This technique was used in pocket watches, which under normal conditions would be in a vertical position in the waistcoat or gilet customarily worn at the time. A tourbillon therefore made perfect sense, because the entire escapement could, for example, complete one revolution around itself per minute. The changing centre of gravity thus had an accelerating effect before slowing down half a minute later in the opposite direction. On average, the existing error was therefore compensated for within one complete revolution. This only works, however, if the watch remains in the same position for at least as long as the tourbillon requires to carry out one complete revolution. It is easy to see that this would rarely be the case with a wristwatch.
Nowadays, with the use of self-compensating alloys - as in Straumann hairsprings - combined with a Glucydur balance, an imperfectly working bimetallic balance is no longer used. The tourbillon can now be used to compensate for the much smaller gravitational error relating to the vibrations of a fl at curve spring. However, this too only works properly if the watch remains in the same position for as long as the tourbillon requires to complete one revolution around itself. As a general rule, we can therefore certainly say that it is better for an error not to occur in the fi rst place than to try and compensate for it afterwards. It is for this reason that we at Moser have developed the Straumann double hairspring escapement system.