2024-08-23

Last modified: 2024-08-24 08:33:54

Clock escapement

Here's another idea for a clock escapement.

Take a normal deadbeat escapement's escape wheel and anchor. Mount a balance wheel behind the anchor, sharing the pivot axis with the anchor (it can be a separate pivot, just on the same axis).

Attach the "stationary" end of the balance spring to the anchor, and constrain the movement range of the anchor.

That's it, you're done!

Will it work? The idea is that as the balance wheel passes centre, it pulls the anchor with it, which unlocks the escapement, and then the impulse is delivered to the balance spring.

I think the drawback might be that not enough of the delivered energy goes into the balance wheel. Probably most of it gets dumped into the banking pins.

Or, if the timing is wrong, it might actually subtract energy with each oscillation, if it is reducing the amount of tension in the spring rather than increasing it.

Gould's "The Marine Chronometer" makes reference to several "constant-force" escapements, which I think my idea would be considered. They all look much more complicated, however, which rather suggests that my idea won't work. It is tempting to 3d print a model anyway.

My prior notes (e.g. 2024-04-28, 2024-05-01) suggest reading Goodrich to remember how to lay out the escapement. The relevant part starts in The Modern Clock by Ward L. Goodrich, Chapter VIII, "The Graham or dead beat escapement", page 109 in my copy.

Here we go:

Now needs a balance wheel on the same axis as the anchor, and a balance spring that fixes to the balance wheel at one end and to the anchor at the other.

And this time we don't even need a mechanism to adjust the "beat" because it works automatically (if it works at all).

With a balance spring:

I don't really like the spring attachment, but it'll do for now.

Just needs something to drive the escape wheel, and a frame.

Maybe the other end of the balance spring should be attached to the anchor. As it stands it won't move very far with each revolution. Although it will move by the same angle, maybe it's equivalent? I think moving through the same angle means it's equivalent.