jes notes Index Gallery . Shaft passers Snap issues

Anniversary Clocks

Last modified: 2024-09-16 21:43:19

Resources

Principles

At first glance the anniversary clock seems pretty much directly isomorphic to a pendulum clock, but it is quite a bit more subtle because of the fact that the "pendulum" and the escapement are joined by a spring. I'm working with a deadbeat escapement, so the following applies in addition to everything that would normally apply to a deadbeat escapement.

While the impulse is being delivered, the part of the spring that the fork is attached to is being driven "ahead" of the position that the balance wheel's rotation would naturally place it. This puts energy in the spring, which accelerates the balance wheel.

If the impulse is completed too quickly, then the fork will "snap back", which will pull the anchor back towards the centre. This effect is unavoidable. The only way the fork will not "snap back" once the impulse is complete is if the impulse is not adding any energy.

If the fork snaps back far enough, then it will pull the anchor back to a position such that the next tooth, due to land on the locking face of its adjacent pallet, will instead land directly on the impulse face. It will then immediately push the anchor back the other way, and the clock will rapidly run down. This is called flutter.

At its core, flutter is caused by the suspension spring "snapping back" too far once the impulse is removed.

Causes of flutter include:

The next problem unique to the anniversary clock is that when the lever and fork are interacting, the effect on the suspension spring is not a pure torque, but is also pushing it to one side or the other. My intuition is that this is a waste of energy, but I haven't proved that.

The effect can be reduced by:

And then the next thing is that there is some minimum amplitude that the clock needs to achieve in order for escapement to occur. You can imagine that clamping the fork near to the suspension spring means that, absent any force from the anchor, it moves in proportion to the balance wheel (if the fork is 1/10 of the way down, then it rotates 1/10 as far as the balance wheel does), and obviously if it doesn't move far enough then it doesn't push the anchor far enough and therefore the escape wheel never escapes.

Increasing the weight on the balance wheel is desirable because of the associated increase in moment of inertia, but the downside is that a balance wheel with a given amount of energy will have a lower amplitude if it has a higher moment of inertia. So if you make the balance wheel too heavy then you won't have enough amplitude for the escapement to work.