A Foucault pendulum shorter than two meters is not so easy to realise. One meter is more difficult, and mine must have this lenght to be one second. Few people have made short Foucault pendulums: I must mention Haym Kruglak and Stanley Steele in 1884 and H Richard Crane in 1884. The imperfections of the system increase when the lenght is reduced, but the worst problem come from the suspension and the device employed to limit the growth of ellipticity. The best way to reduce the ellipticity is to use a Charron ring at one point one tenth of the distance from the top of the pendulum. This hole in the Charron ring must be a highly polished perfect circle.
 

This pendulum is driven by a small coil, wich gives just enough power to keep the pendulum swinging. I've given up the idea of holding the wire in pin wise jaws, as jaws must be perfect and I dont have the tools to make them. The wire go through a very precise hole (+/-1/100 mm!) and is tightened at a point 10 cm further on. This allows setting  the pendulum's lenght (and therefore the time) when swinging without problem. It's not usual to finding  this kind of feature in classical clocks, but it has proved to  work very well. I can even add a bimetalic system wich can push a little the wire when the temperature increases...

And now, lets see what is the difference between this pendulum and the only ones I know wich can also be used like a clock: the two Foucault wallclocks made by H. Richard Crane (University of Michigan). With its clocks, we can read the time at the bottom of the pendulum. At that latitude, it makes a full turn in 18 hours. Therefore Crane stops the rotating pendulum for the required 6 hours every night with a small coil driven by a electronic timer. This clock can be precise, as the timer resets the time every night. The real base of time is not in the pendulum, but inside the timer.
 
 

Some facts: The pendulum is actually driven by a Mumford controller, wich "pushs" the pendulum once every second. The base of time is given by the pendulum's lenght, as usual. Everything is made of brass in order to avoid magnetic troubles. We can read the time (hands for hours, minuts and seconds) on the top and also at the bottom of the pendulum (hours and quarters of hours). Invar is not used, because it is magnetic. I can compensate the temperature's effect on the music wire, on the top of the clock. If  the wire happens to break, the fork halfway stops its fall without damage. No precaution is required when launching the pendulum. The Foucault effect can already be seen after 30 minutes.

Many people still beleive that a Foucault pendulum turns on itself in 24 hours. This is right only at the poles, clockwise at the north pole and counterclockwise at the south pole. The more we go down from the north pole, the more it takes time to make a full circle. In Paris, it takes 32 hours. At the Equator, the pendulum keep swinging the same way without rotation.
 
 

Russians mountains... There is the first test of this clock. It has been taken in my workshop before the pendulum's lenght was set, the Charron ring was not centered and there were many people wandering around, but it shows many interrestings things. This data covers 2 days. We can see that periodic perturbations come back 3 times. If we divide the number of hours of this data with the number of the cycles, we get the number of 16.5 . This is the number of hours the pendulum needs to make a full circle. In fact, it takes the double that time, but the pendulum returns to the same plane of swing in half that.

Perturbations comes from the de-centered Charron ring. If I perform the same test without any Charron ring, I get a flat boring data, an elliptic course and no Foucault effect. It doesnt mean I have to choose between a precise clock or a good Foucault pendulum. I can get both together.

After some days and settings, here is the new clock's data.

 This data shows us that this clock speeds up 7.8 seconds every days, wich is easy to compensate for by lowering the pendulum. Every peak means a complete revolution of the pendulum. Then if we want to set the pendulum's lenght, we have to wait one full revolution (or more) and to set the average. We can also notice the small perturbations at the bottom of every sinusoid, wich reveal a small imperfection in the wire or in the suspension.

This clock is of course unique, but I can make another one wich looks like it if ordered. Please feel free to write me if you are interrested by buying one. The latitude should be given, as the display is personalized. The delay is three month. Special executions are also available if needed.

If you are interrested by the settings of a Foucault pendulum, click here