20 January 2014

An original escapement design?


© Trustees of the British Museum
[source page]

On a recent visit to the British Museum in London, I was interested to see a very curious escapement mechanism in one of the clocks, a creation of Thomas Tompion in the 1670s. I am certainly no expert on clocks, but I had never seen an escapement where the pendulum swings across the plane of the gear wheels.

Frustratingly, the actual mechanism was very small, and in the middle of quite a large glass case, so I could not see it well enough to work out what it was doing. I spent a bit of time thinking about it, and came up with an idea for what might be happening, but now the question was whether I'd got it right.

Turning to the web, even this lovely photo from the British Museum does not quite make the situation clear:


© Trustees of the British Museum [source page]

Some further research lead me to a 1971 pamphlet which contained a drawing detailed enough to understand the mechanism. Conveniently for this post, the drawing is from more or less the same viewpoint as the above photo:

The Tompion Clocks at Greenwich and the Dead-Beat Escapement, Derek Howse, (originally Antiquarian Horology, December 1970 and March 1971), p.29, drawing from Donald Cufflin, National Maritime Museum. Annotations 'P' and 'W' added.

I had got it nearly right.

The simplified animation below shows the difference. On the left is how I imagined the mechanism, and on the right is the actual mechanism. In each half: The oscillating pair of long black pieces is the forked pallets ('P' in the figure above, for the actual mechanism), to be imagined as connected to the swinging pendulum. The central vertical grey band represents an edge-on view of the wheel with radial pins ('W' in the figure), as if driven up the screen by the weights.

I had guessed that the pins were staggered side-to-side, but in fact the pallets were staggered front-to-back. The impulses are provided when the pin slides along the oblique faces at the pallets' ends.

Implementing my version

I thought my mechanism ought to work, and so I made it out of Lego. It is different from the real one in a couple of ways, besides the main difference of the pin and pallet arrangement:

  • I could not easily fit more than four pins round the wheel, contributing to the clunkiness of the result. The real clock has thirty pins on the pendulum wheel.
  • Compared to the real clock, the interaction between the pins and the pallets is offset a quarter of the way round the wheel. This made it easier to attach the rounded-ended pieces forming the pallets, and it also keeps the impulse force in the plane of the pendulum.
  • My pallet arms would need to be curved to make the mechanism truly dead-beat. There is a very slight recoil because they are straight.
  • The real pendulum is 13 feet long. Mine is not.

But it does work! This video is a composite of three shots, taken consecutively from different viewpoints:

Comments are closed.