Silicon is nowadays well known and not new in business but how to produce and where the little details are for gaining 96 hours of power reserve - this is what HORAGE is really good at.

What are the main reasons for the significantly higher efficiency of the silicon escapement compared to the traditional Swiss lever escapement? Let's start with the weight. Steel, which is usually used in the traditional escapement, has an average density of 7.85 g/cm³, whereas silicon is only 2.33 g/cm³ and therefore 3.3 x lighter. This means that when the escapement is triggered, in our case with a frequency of 25,200 beats per hour, the escapement is triggered over half a million times per day. So significantly less energy is consumed to move the silicon anchor.

Second point. The precise manufacturing tolerances of both the milled parts (escapement positions) and the silicon components make it possible to set very small safety margins (tolerances) and thus minimize force-displacement losses and thus make them more efficient.
Third point. Silicon is anti-magnetic, hence the name of our first model with the manufacture calibre K3, DecaFlux. This means that other components are less likely to disturb or influence the system. In the watch industry, isochronism is said not to be disturbed by magnetism.
However, I would also like to point out the challenges in relation to silicon components. The handling of the separation from the wafer as well as the subsequent operation and assembly requires a high level of skill and a stable geometry of the escapement components, as we are talking about the brittleness of glass.

The myth that silicon escapements do not need lubrication should also be dispelled; they simply need much less.
What advantages can be expected from a silicon hairspring? Well, as with the escapement, the weight of the hairsprings is decisive. Due to the low dead weight, there is no additional imbalance of the hairspring during one-sided expansion, which would lead to unfavourable rate values in the suspended positions of the watch. Furthermore, the anti-magnetic advantages of the hairspring are even greater, as magnetization would have serious consequences. Most of you are probably familiar with this problem because you have come too close to a fridge or handbag magnet. The result is a watch which may runs several minutes a day fast.
Another interesting point is the coefficient of thermal expansion of silicon. This is very low. Anyone familiar with chronometer certification will know how crucial this value is for passing the certification, as the watch is tested at 8°, 23° and 38° at the COSC (Chronometer Testing Centre in Biel) using appropriate criteria. While Benjamin Franklin was still chasing after the phenomena of nature and the predictions of temperatures. Thomas Earnshaw tried to equalize the temperature fluctuations of the weather with the compensation balance wheels. Today's balance wheels (French balancier) are no longer balance wheels with two materials that expand differently and thus compensate for each other. Modern alloys (material fusions) are used that have the smallest possible coefficient of thermal expansion.
With traditionally rolled steel hairsprings, very constant spiral thicknesses can be produced, which in turn means a constant frequency. With silicon hairsprings, on the other hand, the complicated manufacturing process has many influences that affect the frequency and cause a wide variation. This makes it much more difficult to match the hairspring to the balance wheel than with conventional steel hairsprings.
However, this alone does not provide a power reserve of an astonishing 96 hours (4 days). Let's talk about the other small optimizations we made during our experience with K1, K2, KT1, KT2 and let us not forget the KTM.
Let's start with the barrel and the customized mainspring. Thanks to the more efficient escapement parts made of silicon, we can use a thinner and longer mainspring that delivers less force but more driving hours to maintain the same amplitude (amplitude of oscillation) on the balance wheel. The Nivaflex Plus material that we use for the mainspring is precisely adapted to our required torque, so that we can count on a constant force in production. Otherwise achieving COSC levels would be a little tricky.

Another crucial point is the tooth profiles of our gears, which we calculate in-house. In theory, many things are possible, but the reality is a different story. Manufacturing and assembly processes require a certain amount of leeway (tolerances) to function almost smoothly as an assembled gear. The aim is to minimize this friction as much as possible. Not only the tooth geometry plays a decisive role here, but also the coating, which brings us to the next exciting topic.
Many movements on the market have hardened steel pins as gear bearings. Why? Because the lateral bearing pressures are high, and they must therefore be robust. This is a little different with us. The thinner and longer mainspring has significantly less force, which is also advantageous for manual winding and automatic winding. This is also the reason why our movements wind automatically so efficiently. Due to the lower force, lateral bearing pressures are lower, which allows us to mill the bearing pins and save valuable time during assembly. Hardness is not the only reason; we can also replace the deliberately low friction of polished steel pins with modern coating processes and thus also new materials, and so well that wear tests show no differences.

I am a watchmaker who probably had one of the most traditional and value-laden training programs in Switzerland and therefore have a certain awareness of historical watchmaking. BUT, the possibilities we have today are great and HORAGE uses them to be innovative and develop great timepieces for you.
Read about our latest release powered by the K3 calibre the DecaFlux HERE
See you next time at HORAGEOLOGY
All the best, your watchmaker Lenny.
Hmm, exciting Fieldwalker. If I may ask, what were the arguments of those who described silicon as not “pure”? Good to hear that i was able to deliver some nice counterarguments :)
Haha thanks, we really appreciate the compliment^^ Unfortunately, there are many costs here that we cannot influence, such as the cost of the individual wafer or the cost of a new exposure mask for the silicon parts.
But of course this is always an important part for us to continue to develop and produce cool watches for all Joe Lunchbox / Sally Housecoat types ;)
Great article. A few questions on my end:
How good is the rate stability on both ends of the power reserve?
Curious how you landed on the 3.5Hz frequency for all Horage movements. Is it the sweep spot for power reserve vs accuracy vs ability to design compact movements?
Do you have measurements that compare Horage's movements to the mainstream ones from brands like ETA, Sellita, Rolex, Grand Seiko? From an efficiency, accuracy, etc perspective? SJX had an interesting article where they tried to compare movement through a quantified measure (HDF)
Keep up the good work!
JF
Brilliant write up Lenny. It's a pleasure to read, thank you. So many advantages of Si.
Thermal conductivity sometimes comes across my mind when picking a watch (Yay for Ti and Ceramic! Boo sss for steel!) but thermal expansion never crosses my mind. The joys of modern watches that they are mostly free from this issue.
The longer less powerful mainspring aspect and its implications for the gear train are fascinating.
If it's ok, I've a bunch of questions:
I've heard some suggest that a silicon hairspring is not tunable or fixable, and therefore inferior. It can only be swapped out, and that can be difficult to get right. What's the counter point to this view? I'm guessing it's an outdated argument?
UN does a pretty amazing dial using silicon marquetry - making a spectacular blue kaleidoscope of sorts. Does Horage's expertise with Si give you the skills to do your own version of a dial featuring the material? Can you try? 🙂😉
Does the longer less powerful mainspring result in a flatter more optimized torque curve for the gear train? ( i.e. relatively less fast at full wind, less slow when near empty.)
Thanks buddy!