© text: L. Oechslin, translated by M. Hanke, pics: MIH, 2002. 2006

In 2001, the Centre Suisse d’Electronique et de Microtechnique SA (CSEM) contacted me, which led to a first meeting at Neuchatel on November 8, 2001. The reason for contacting me was their wish to know whether there were watch parts other than those just manufactured for the “Freak” escapement, which could possibly be made from silicon.

The main goal in the development of the “Freak” escapement wheels was to make them as light and tough as possible, since otherwise, the advantages of the escapement system would have been nullified by the larger inertia of two wheels’ mass. Since silicon is so light and stable, it was the material of choice for compensating the inertia problem. It is possible to build up silicon to very thin layers, and it was CSEM that developed a photolithographic method to retrieve flat parts from these layers, keeping the loss of precision at an absolute minimum.

Development of this method was very expensive, which is why CSEM searched for additional possibilities to commercialize their silicon technology.

After the first conversation, I suggested them to initiate trials which should prove if silicon with its natural toughness and partial flexibility would be a good material for balance springs in watch movements. Due to Nivarox’ monopoly in the production of balance springs for the watch industry, their exists a firm demand for equal or even better alternatives.

A new spring would need to possess the same technical capabilities as Nivarox:

1. It does not change its length due to temperature changes, and

2. its flexibility is not affected by temperature changes as well, which helps keeping accuracy changes of watches due to temperature differences as small as possible.

Until now Invar and Elinvar were materials fulfilling these conditions.

Besides that, however, I found that producing the spring from silicon, would have two further, decisive advantages:

3. The possibility to vary the spring’s thickness makes the use of a specific end curve unnecessary, and allows the spring to “breathe” centrally, and

4. the collet and stud could be integrated into the design of the spring’s inner and outer ends, thus simplifiying its integration into the watch.

Following these conditions, I designed a first spring, which was produced by CSEM in several different sizes, and mounted into a pocket watch here at the MIH, in order to check its capacities in action.

The test protocol found

a) the basic practicability of silicon springs for watch escapements, even if the effect could not yet be achieved to an ideal extent;

b) the correctness of the idea to achieve a centric pulsement of the spring by means of variations in thickness, and

c) that the mounting of the spring into the watch movement could be drastically simplified by appropriate end shapes.

However, it was also apparent that the silicon, as manufactured by CSEM, still had not the necessary physical capabilities to make it unaffected by temperature changes.

It has been assumed that this stability could be achieved by an appropriate coating of the the basic silicon material.

In any case, these trials initiated further surveys and experiments to intensively search for Nivarox alternatives.

(Read the detailed test report and protocol here)