The SeeSaw Gravity Escapement part 3
by Roger Bunce

Adjusting the escapement

With the pendulum rod vertical and still, and the weight supports in the up position, the top surfaces of the arms were arranged to be below the weights by half the impulse distance. Hence, when the pendulum swings past the vertical, the arms lift the weights. The weight support stops were adjusted, so that with the weight supports in the down position, the pendulum arms lifted the weights clear of the weight supports, again, by half the impulse distance (the full impulse distance being about 0.5 to 1.0mm). In other words, impulse is symmetrical about the pendulum's vertical position.

This satisfies what is known as 'Airy's Condition'. Sir George Biddel Airy was an English 19th century mathematician and Astronomer Royal. He showed that this condition negates escapement error. A further implication is that this minimizes timing deviation due to changes in impulse. Of course, the whole idea of this escapement is that the impulse does not change. Nevertheless, as an added precaution, the escapement may be set to satisfy Airy's Condition.

It will be appreciated that as a pendulum arm lifts its weight clear of a weight support, the other arm lowers its weight onto its weight support. This weight support is then reset, to its original height, as the escape wheel indexes, and the resetting roller turns the reset lever.

However, there is always a slight delay before the prop resets the weight support and weight thereon. This occurs as the prop turns and the unlocking roller rides up the escape wheel tooth. Further delay occurs due to clearance between the reset roller and reset lever. Hence, the pendulum arm is below the weight (has released the weight), before the weight support raises the weight. So the pendulum is quite oblivious to the fact that the weight has been raised (reset).

Alternatively, the weight support stops may be set just a little higher (say 0.1mm), than the optimum Airy requirement. This further ensures that resetting only begins when the pendulum arms are below the weights.     

The weight supports are fitted with counterweights. These are adjusted so that, with the weights lifted clear of the weight supports, and with the props in the lower position, the distal ends of the weight supports lower against the stops. The pivotal action must be reliable, but not too forceful, since excess force interferers with the unlocking action of the props.

The prop counterweights are adjusted so that, with a weight resting on a weight support, there is insufficient torque on the prop to overcome the frictional resistance between the roller clutch and cam of the weight support, to allow the ball bearing roller of the locking arm, to roll off a tooth of an escape wheel. It will be realized that in this condition, the frictional resistance between roller clutches and cams is greatly increased because the roller clutches cannot turn. Hence, the escape wheel remains locked.

Conversely, when a weight is lifted clear of its weight support, the frictional resistance between the roller clutch and cam is reduced. The torque applied to the prop, via its counterweight, is then sufficient to cause the prop to pivot. Hence, the roller of the locking arm rolls up the flank of an escape wheel tooth, allowing the tooth to 'escape', and the escape wheel assembly to index.

It will be appreciated that during 'unlocking', the periphery of roller clutch does not roll, but slides against the cam. However, as mentioned earlier, without the weight in place, the force between cam and roller is minimal. The torque provided to drive the escape wheel must be sufficient to reliably index the escape wheel once within half the cycle time of the pendulum, but not enough to prevent the roller of the locking arm from unlocking the escape wheel.


Operating sequence (first impulse mode)

Figures 15a – d show the operating sequence of the SGE in which the pendulum is impulsed first in one direction and then the other.

The oscillation is begun by displacing and releasing the pendulum manually. The oscillation then continues as long as the drive means in active. Figure 15a shows the pendulum swinging to the right from the centre position. As soon as the arm of the pendulum begins to lift the RH weight, the frictional resistance between the roller and cam of the weight support is reduced, and the prop pivots clockwise against its stop. The RH weight support lowers and is restrained by its stop. Meanwhile, the weight continues to be lifted by the arm of the pendulum.

It will be appreciated that as the weight is lifted clear of the weight support, the weight and weight support begin to move in opposite directions as described earlier (Prop). Hence, any friction at the weight support pivot, or the weight of the weight support itself, has no affect whatever on the pendulum.  Furthermore, the action of lifting the weights out of the Vees of the weight supports does not involve sliding or rolling friction, and the weights self-align, so this action also has no affect whatever on the pendulum.

There is no need to provide lubrication between the arms, weights, and weight support. Indeed, this is highly undesirable since it would have an adverse affect on the timekeeping, due to the variability of surface tension forces.

Concurrent with prop pivoting clockwise, the distal end of the locking arm lifts, causing the escape wheel assembly to unlock and index clockwise. The LH resetting roller of the escape wheel assembly turns the LH resetting arm clockwise. This causes the LH prop to become vertical, and the locking arm to lower, thus locking the escape wheel. Also, the LH prop raises its weight support and weight thereon. It will be appreciated that when this happens the LH pendulum arm is lower than the weight. Hence, raising the LH weight has no affect whatever on the pendulum, as described above (Adjusting the escapement). At this stage in the sequence, the positions of the components are shown in Figure 15b.

Meanwhile, the pendulum continues to swing to the right. After reaching its maximum amplitude, it begins to swing left. The distal end of RH pendulum arm lowers and places its weight in the Vees of its weight support. As the pendulum continues to swing left, the pendulum arm continues to lower.

The difference between the height, when RH arm begins to lift weight from the weight support, and the height when it replaces the weight back onto the weight support (now in the low position), provides the impulse to the pendulum for the first half of its cycle, as described above (Impulse and reset). The impulse for the second half of the cycle is provided by the LH weight.

The pendulum continues to swing left so that the pendulum arm lifts the LH weight clear of its weight support. The frictional resistance between roller and cam of the weight support is reduced, and the prop pivots anticlockwise against its stop. The weight support is now in the low position - supported by its stop, and is ready to allow the LH weight to impulse the pendulum for the other half of the cycle. Concurrently, the escape wheel is unlocked and the RH resetting roller of the escape wheel assembly turns the RH resetting arm anticlockwise to raise the RH prop to the vertical position which, in turn, raises weight support and weight thereon. At this stage in the sequence, the positions of the components are shown in Figure 15c.

The pendulum continues to swing left until it reaches the extremity of its swing, and then it begins to swing right. As the pendulum reaches the centre position, it will be noted that Figures 15a and 15d are identical (except that the escape wheel has indexed). In other words, the pendulum has completed one oscillatory cycle. The escape wheel continues to index as the pendulum continues to swing, and all the other components move cyclically according to the aforementioned sequence.


Go to Part one

Go to Part two

Go to Part Four

Figure 15a. SGE with the pendulum in the centre position and swinging to the right.


Figure 15b. SGE - the pendulum having swung right.


Figure 15c. SGE - the pendulum having swung left.


Figure 15d. SGE - the pendulum having swung back to the centre.

WATCH IT RUN!

http://www.youtube.com/watch?v=A1yH9vJN__whttp://www.youtube.com/watch?v=A1yH9vJN__wshapeimage_3_link_0