Steyr
132
Well, this is pure application of physics law from Shannon-Nyquist. It is usually taught around 9th grade
Apr 06, 2022,17:01 PM
This is related to discretization (going from one analog event to a numerated result). That is, for instance, the reason why CDs are sampled at ~44KHz while the ear is limited to ~22KHz. You need a reference that is twice faster than the measurement instrument. Otherwise you fall into the measurements uncertainties as defined by the BIPM.
Let's take for instance 4 different events. Each during exactly 1.98s ; 1.89s ; 1.82s ; 1.62s. The measure at a tenth of a second should give us respectively : 2.0 ; 1.9 ; 1.8 ; 1.6. Or more precisely, values a value in between 2.08<1.88 ; 1.99<1.79 ; 1.92<1.72 ; 1.72<1.52. A measure that is not within the range is false.
Now, take a look at what happens depending o, the frequency of the watch :
A 5Hz watch will not be able differentiate from a 1.62 to 1.98 duration. Two measures are false, two are right.
10Hz will give correct measures with these examples, but my choices are not relevant enough here. A 1.99 duration with a trart on the first beat would have confirmed the rule for instance.
20 Hz will allow us to reduce the uncertainty to less than a tenth. That is exactly the point. Therefore, Shannon Nyquist theory applied to a watch measurement is demonstrated.
You can also raise the other errors a mechanical chronograph introduces. such as the Beat error on the escapement. Not relevant except on the Tag Micrographer. This is also the reason why one should take the oscillation not the alternance.
The horizontal clutch is also bringing a noticeable error like on a El-Primero, even with a 360 teeth chronograph wheel. This only is already representing a 0.4s uncertainty on this mechanism.
Conclusion, this Patek is not capable to measure with a 1/10s accuracy. Barely, 2/5th of a second.