I am (as you may know) a bit of a watch geek.
Many of my watches are self-winding, and the mechanism is both simple and elegant. A single rotor (effectively, a half-disc) swings two and fro as you move. A beautifully simple gear mechanism converts this movement - in _either_ direction - into winding the mainspring.
It's a very nice system (watchmakers were harvesting "wasted" energy centuries before we thought of attaching generators to hamster wheels), but not perfect. Because of the large reduction in gearing needed to wind the mainspring, and because of the need to change the gearing between forward and backward turns of the rotor, there is some lost motion in the system.
Some arm movements will set the rotor spinning, in which case the lost motion is negligible. But many other movements just swing the rotor back and forth, with considerable losses at each swing. If the rotor were of a different diameter, it would be set spinning at different frequencies, but would still have "dead spots" where it just swings back and forth.
What is needed, therefore, is a multi-rotor system.
Instead of a single rotor (which, as mentioned, is like a half-disc, weighted around its edge), we have two or more concentric rotors. Each has a different diameter (again, with its edge weighted), and they fit together like a series of nested jar-lids. Crucially, the different diameters will be set spinning by different arm movements.
The clutch/gear system necessary to capture energy from whichever rotor is turning the most will be a little complex, but not unduly so. In fact, it can be based on the same principle as the existing single-rotor, direction- rectifying clutch.
The result will be more efficient energy capture. This, of course, is of little use on a regular watch (which is adequately powered by normal self-winding mechanisms), but becomes more important for watches with greater numbers of complications which, at present, cannot be adequately maintained by a simple self-winding mechanism. Also, given that people will spend extra thousands of pounds for a watch with a largely decorative tri-axial tourbillon, a multi-rotor self-winder ought to be worth a fair bit.-- MaxwellBuchanan, Feb 02 2016 It seems to me that the energy generated is proportional to the off center mass of the flywheel. While dividing up the mass between 3 different flywheels with different resonant frequencies might help some (15 - 40% maybe?), if we could increase the mass, we would get a linear increase in energy.
Assuming we don't want to change the overall size or weight of the watch, what we need to do is take advantage of the existing mass. I can't see any way to use the mass of the case, watch face, and crystal which by convention are expected to remain stationary relative to the wrist, but everything else is fair game. Therefore the gearing mechanism needs to be miniaturized to fit in a little less than 1/2 the volume, but then constructed out of depleted uranium the bring the weight back to where it started. Have that mounted off center so that most of the clockwork can swing around inside the case that is approximately half empty. Some clever gearing will be necessary to keep the hands steady and interface with the stationary knob, but that would just be some variations on a differential gear.-- scad mientist, Feb 02 2016 //the energy generated is proportional to the off center mass of the flywheel.//
Yesish, but a proportion of the motion is lost. In a simple but well-made rotor (with bidirectional gearing), I would guess that something like 20 degrees of motion is lost. Hence, if the rotor is swinging back and forth through a 120 degree arc, only 80 degrees of that movement will be harnessed.
In contrast, if you can get the rotor swinging, you lose 20 degrees initially, but then all further rotation is available for winding.-- MaxwellBuchanan, Feb 02 2016 Perhaps a combined blood pressure cuff and watch, built with hydraulic powered means to measure time and pressures. Thump bump about once a second.
Sales would increase if your Doctor had to prescribe the watch for you.
"But it stopped short — never to go again — When the old man died."-- popbottle, Feb 02 2016 // we have two or more concentric rotors.// What about 3 orthogonal axes for rotating masses, winding a single spring ? Any reciprocating, circular motion would add energy to the system, and linear accelerations too, to some extent. Not suitable for a watch, but maybe usable for other applications-- piluso, Feb 02 2016 Orthogonal rotors are an excellent idea. And, with the current fashion to make watches as big as possible, probably feasible.-- MaxwellBuchanan, Feb 03 2016 I would like to see this idea applied first to long-case grandfather clocks on ships, which could easily be made to recover energy from the rocking and pitching motion of the waves. This would save money for the operators of ocean liners who would no longer have to clock-winders.-- hippo, Feb 03 2016 random, halfbakery