Pestoline (technical)

Because a Foucault pendulum is only easy to make if it is longer than 2 metres. But mine must have a balance of 1 metre to be able to indicate the seconds. Few people in the world have been able to make small pendulums that work satisfactorily. I must mention here the pioneers Haym Kruglak and Stanley Steele in 1984 and H Richard Crane in 1981. The two main problems with small pendulums are suspension and precession of ellipticity. The problem of suspension is by far the most important, as it has to be perfect or the pendulum will only swing in the plane that is easiest for it. Another solution is to cast the wire in metal. But the melting temperature often destroys the specifications of the suspension wire. A more common suspension method is to run the wire through a precision chuck. But here too the pendulum will choose a preferred position and stabilise itself on a plane. Commercial chucks all have small imperfections. Richard Crane used to run his in a lathe to polish the jaws with a very fine abrasive. Only then could the pendulum work. The other problem is how to eliminate the precession of ellipticity. This phenomenon only occurs in small pendulums and gets worse as the length is reduced. This undesirable effect can be somewhat reduced by interposing a perfectly machined and polished Charron ring at the upper tenth of the length of the pendulum.

This was the first prototype that really worked, the one that was intended to serve as a test bench for the construction of the others. It was also destined, once all the tests were completed, to end its days in my kitchen. Because I didn’t have a clock at home….

The bracket has three functions: to support the balance wheel, to guide the wire of the balance wheel very precisely (to a hundredth of a millimetre) and to allow the length of the balance wheel to be adjusted even while it is moving. This is a function that is hardly ever found on clocks, but which should prove very useful in the future. I abandoned the idea of a chuck because it was too difficult to machine with my poor resources and did not allow any adjustment once clamped. Instead of being clamped between jaws, the steel wire (0.18mm) passes through the hole of a sapphire, climbs around a wheel, and comes down to be blocked at the bottom. An adjustment screw allows to bend the wire after the wheel and thus to act on the height of the balance. I can even add a bimetallic temperature compensation if I want.

Now let’s see what will differentiate this pendulum from all the others that have been made so far, namely that it also serves as a clock. As far as I know, only one other man has ever attempted this: H. Richard Crane (University of Michigan). The passing of time was read by reading the passage of the pendulum on a graduated circle at the bottom. But his pendulum cannot be considered a real clock in the sense that he needed an external time base, a timer that made the pendulum stop rotating 6 hours a day. Mine, on the other hand, has to do everything at once: real clock and Foucault pendulum. From then on, the problems raised become very interesting. For example: how can I ensure the temperature compensation of the pendulum when I am not allowed to use invar (as invar is magnetic, this pendulum would quickly become a compass) or how can I indicate the time with the pendulum knowing that it will have a scale incompatible with 24 hours?

Many people still think that a Foucault pendulum rotates on itself every 24 hours. This is only true at the North Pole clockwise and at the South Pole counter-clockwise. The lower the latitude, the slower it rotates. In Paris, it is 32 hours. At the equator, it has a theoretically infinite rotation time. Crane stopped the revolution of the balance wheel 6 hours a night (between midnight and 6 a.m.) by means of an additional electromagnet controlled by a clock.

Some facts.
Pendulum electromagnetically propelled by an electromagnet whenever it detects the passage of the pendulum. Brass for almost all parts to eliminate magnetism problems. Time display on the dial (hours, minutes and seconds) and time display at the end of the balance wheel. The disc in the upper third serves to block the fall of the balance wheel should the wire break.

The stabilisation of the amplitude uses another discovery of Léon Foucault: the brake. This is a non-magnetic graduated ring, with very precise dimensions, placed at the end of the balance wheel’s stroke. The passage of the magnet causes induced currents known as “Foucault currents” which brake it at each oscillation.

A roller coaster: that’s what the first Pestoline test was like. This test was taken in my workshop while the height of the balance is not adjusted, while the Charron ring is not centred and while I work all day around it. Nevertheless, one can discover quite a few interesting phenomena. On this document, which covers two days, we can see that cyclic disturbances occur three times. Dividing the number of days covered by this graph by the number of cycles gives us the figure of 16.5: that’s the number of hours it takes for the pendulum to make a complete turn on itself. In fact, it takes twice as long, but as it passes the same place twice, we can divide by two. The disturbances we can see come from the Charron ring. It is not yet well centred and we can see some machining defects. I did the same test without the Charron ring: the line was perfectly flat, but the pendulum stroke was elliptical and the eddy effect was barely detectable.

After some adjustments, centering and balancing, here is what the computer reveals:

This graph shows us that Pestoline is advancing by 7.8 seconds per day, which is easily adjusted by lowering the pendulum a little. But more importantly, the succession of peaks (each indicating a revolution of the pendulum) has become more precise. We can also see recurrent disturbances at the bottom of the curve, which prove that there is still a mechanical anisotropy in the wire or the suspension. This is why this clock serves as a test bench for the improvement of the next one. Because the choice to be made with this type of clock is not the accuracy of the time or that of the eddy current effect. Pestoline has proven that both can achieve a good result.

(July 2003 ) Pestoline was awarded first prize in the Kinetic Art Organization’s “ingeneering ingenuity
“
competition.

Those who own Bryan Mumford’s MicroSet software can download it on request. They will be able to find out what happens in a millionth of a second in the heart of an eddy pendulum, the effect of disturbances caused by air currents, parasitic oscillations, etc. This sample covers 5 days of testing

HOW TO SET UP A FOUCAULT PENDULUM ?

It is almost a case of solving the equation with 6 unknowns: The level of the coil, the centring of the coil, the level of the suspension, the centring of the wire attachment, the anisotropy of the suspension wire and the centring of the Charron ring. Change just one of these parameters, and the pendulum will only swing without rotating. This method is therefore a logical guide to getting the pendulum to work. But let’s always keep in mind that a working pendulum is almost a miracle. Setting up a pendulum takes time and a lot of logic. Make a note of all the changes you make each time. Never rely on luck or chance, because they will waste a lot of time and teach you nothing. A pendulum can be tamed. Above all, note the position of the oscillation plane rather than the time, as this can teach us a lot more.

METHOD FOR ADJUSTING A FOUCAULT PENDULUM?

It will almost be a question of solving an equation with 6 unknown factors: The level of the drive coil, centering of the drive coil, the level of the suspension, the centering of the attachment of the suspension wire, anisotropy of the suspension wire and the centering of the Charron ring. Change only one of these parameters and the pendulum will do nothing but oscillate in one plane without turning. This method is thus a logical guide to arrive so that the pendulum functions correctly. But always keep in mind that a Foucault pendulum that works is nearly a miracle. The adjustment of a clock pendulum takes time and good of logical sense. Note each time all the modifications that you make. Never count on chance or luck, because they will make you lose much time and you will learn nothing. A pendulum can be tamed down. Note especially the position of the plane of oscillation more than with respect to the hour, because it can teach us much more.

1) Time needed for the adjustments

It depends very much on the latitude where the pendulum is located. A pendulum located in Singapore needs 1070 hours to make a complete revolution: the time it takes each time you make an adjustment… At the latitude of Paris, it takes 30 hours to make a complete revolution. You can divide this time by two because the pendulum will go back the same way after 15 hours. For rough adjustments, wait one hour before seeing the results. For final adjustments, wait 15 hours. Allow a minimum of two weeks at Paris latitude.

1) Time necessary for the adjustments

It depends much on the latitude where the pendulum is. A pendulum located in Singapore needs 1070 hours to make a complete revolution: the wait time each time you carry out an unspecified adjustment…So the time for the settings under the latitude of Singapore can reach 6 month. At the latitude of Paris, it takes 30 hours to make a full rotation. You can divide this time by two because the pendulum will pass by again by the same way after 15 hours. For the coarse adjustments, wait one hour before seeing the results. For the final adjustments, wait 15 hours. Thus count on two weeks minimum at the latitude of Paris.

First operations

1) Adjust the general level of the pendulum (bubble level
)2) Fix the pendulum firmly against a wall. It is very important that the pendulum is perfectly stable.
3) Check the level on the wire at the height of the Charron ring (stabilised pendulum, using a cone or dynamic pendulum using an oscilloscope).
4) Adjust the level of the electromagnet with the spirit level.
5) Visually adjust the axis of the coil in relation to the magnet.
6) Carefully launch the pendulum

If all these operations are done in order, the pendulum should not work. You will have a better chance of getting a ballpoint pen to balance on its tip the first time..

The first operations

1) Set the general level of the pendulum (spirit level
)2) Firmly fix the pendulum against a wall. It is very important that the unit is perfectly stable.
3) Adjust the level where the wire passes through the Charron ring. (stabilized pendulum, using a cone or dynamic clock pendulum using an oscilloscope)
4) Regulate the level of the electromagnet with the spirit level.
5) Visually align the axis of the drive coil compared to the magnet.
6) Launch the pendulum delicately

If all these operations are carried out in order, the pendulum should not function. You will be more likely to make a ballpoint pen balance on its point at the first attempt…

2) Fine tuning

1) Formalities for electrically adjusting the axes and levels of the coilIMPORTANT
: First disconnect the coil from the oscillating circuit.- Allow the pendulum to stabilise until it is completely still.- Send a few small one-second pulses using a 9-volt battery, respecting the polarities: the pendulum must be driven by the coil and not attracted.- Note the direction of the oscillation. – Let the pendulum stabilise again. At this point we still do not know whether this movement is due to a wrong centering of the coil or whether it is not perfectly level, as both errors cause the same effects.- Note the location, then move the coil a little in the direction of the first oscillation.- Repeat the operation as often as necessary until the pendulum is stable.- Now turn the whole suspension half a turn.- Feed the coil again. The pendulum must remain stable. If not, it means that the electromagnet is not level, that the axis of the pendulum is not perfect or that the magnet does not have a constant magnetic field on its surface. Note that a bent magnet and a tilted coil can make you think that everything is fine because their effects cancel each other out!
Restart the pendulum. It still doesn’t work? don’t worry, it’s normal.

2) Precision adjustments

1) Electrical formalities to regulate axes and levels of the drive coil:

– IMPORTANT: Initially disconnect the drive coil from the oscillating circuit.
– Let the pendulum stabilize until it is completely motionless.
– Send some small one second impulses by using a 9 volt battery, respecting the polarity: the pendulum must be driven out by the drive coil and not attracted.
– Note the direction of the oscillation.
– Again let the pendulum stabilize. At this point, we still do not know if this movement comes from a bad centering of the drive coil or if it is not perfectly level, because these two errors cause the same effect.
– To adjust, move the drive coil a little in the direction of the first oscillation.
– Repeat the operation as often as necessary until the pendulum is stable.
– Now turn the whole suspension a half-turn.
– Feed the drive coil again. The pendulum must remain stable. If not, that means that the electromagnet is not level, that the axis of the pendulum is not perfect or that the magnet does not have a constant magnetic field on its surface. Remaking the tests, you will arrive finally has to regulate axes and levels correctly. Note how one twisted magnet added with a leaning drive coil can make believe that all is well because their effects are cancelled!
Start the pendulum again. It still does not function? Rest assured: it is normal.

3) Locate the problem
Is it at the top (suspension), at the bottom (solenoid) or in the wire? To find out, simply change one of the parameters slightly. It is best to start at the top.

3) Localize the problem
Is it at the top (suspension), below (electromagnet) or in the wire? To find out, it is enough to slightly modify one of the parameters. It is preferable to start with the top.

4) Symptoms and solutions –
The pendulum always oscillates in the plane where it was sent, no matter how silky it is.
This is the worst case scenario.

4) Symptoms and solutions
– The pendulum always oscillates in the plane where it was started, no matter wich one.
It is the worst case of figure.

– The pendulum starts well, makes an ellipse and stopsCentring
and/or rough levels of the electromagnet or suspension. Review the entire procedure #1.

– The pendulum starts well, makes an ellipse and stops.
Centering and/or coarse levels of the electromagnet or the suspension. Re-examine all of procedure n°1

– The pendulum makes an ellipse and goes in any direction?
That’s better, but there’s still a way to go. The problem is probably at the bottom.

– Does the pendulum make an ellipse and turns in either direction?
It is already better, but there is still way to go. The problem is probably at the bottom of the pendulum.

– The pendulum always stays in the same position.
It is likely that one of the levels is not right. To confirm this, turn the whole suspension in one direction and wait to see if the pendulum direction follows the movement a few hours later. If nothing happens, the problem is at the bottom. Turn the solenoid a quarter turn and see what happens. By habit, I always turn the suspension about 1/4 hour clockwise.

– The pendulum always remains in the same position.
It is probable that one of the levels is not right. To confirm it, turn the whole of the suspension in a direction and wait to see whether the direction of the pendulum follows the movement later in a few hours. If nothing occurs, the problem is at the bottom. Turn the electromagnet a quarter of turn and note what occurs.

– The pendulum starts in the right direction, but stops after a few hours.
Things are finally starting to get interesting. From now on, we will have to follow an order to find the causes of the problem. First of all, if you don’t want to waste too much time, all changes should be made clockwise for the northern hemisphere and counter-clockwise for the southern hemisphere. You should always follow the natural movement. If you go against it, it will take much longer to see the effects of your changes.

– The pendulum starts in the good direction, but remains at the end of a few hours, always in the same plane of oscillation.
Things finally start to become interesting. From now, it will be necessary to respect an order to find the causes of the problem. First of all, if one does not want to waste too much time, all the modifications must be done clockwise for the northern hemisphere and anticlockwise for the southern hemisphere. The natural movement should be always followed, because if you go against it, it will take much more time to note the effects of your modifications.

The problem is probably at the top. To find out, turn the whole suspension a quarter turn after marking its position. The pendulum should follow the movement and stabilise a few hours later a quarter turn further back. If this is the case, it is proof that the problem is at the top and not at the bottom. Mark the new position and turn the suspension a quarter turn further.

The problem is probably at the top. To find the problem, turn the whole of the suspension of a quarter of turn after having marked its position. The pendulum should follow the movement and be stabilized a few hours later a quarter of turn further. If it is that, it is the proof that the problem is at top and not at the bottom. Mark the new position and turn the suspension further a quarter of turn.

Here it is possible that the pendulum will start to work properly. This would mean that there are two different faults whose effects cancel each other out : if the wire has internal tension in one direction and the suspension is not perfectly level, it is possible that a state of equilibrium is reached and that these two forces cancel each other out. In this case, the adjustments made will not be valid for another wire.
If the direction of the pendulum always follows the position of the suspension whatever it is, then the problem(s) is at the top. It could be the suspension, the wire, the centring of the wire or the attachment of the wire. To find out if it is the wire, turn it a quarter turn with the suspension fixed. If the pendulum follows the movement, the problem is with the wire or the wire attachment. It may be that the wire attachment is a little off-centre.
To find out if this is a suspension fault, turn only the thread guide. If the pendulum follows the movement, we know the culprit. In this case, the last sapphire crystal is chipped or badly set. To find out if it is the wire attachment, turn it while keeping the whole suspension fixed. If nothing happens, the problem is at the bottom. It is important to know that an electromagnet whose level is off by even 0.2 millimetres will cause the pendulum to stall. The first pendulum I made could only work if a small sheet of paper was slipped under the coil at a very precise point. This is to give an idea of the importance of levels.

Here, it is possible that the pendulum starts to function correctly. That would like to say that there are two different defects of which the effects are cancelled. Indeed, if the wire has an internal tension in a direction and that the suspension is not perfectly level, it may be that one arrives in a state of balance and that these two forces are neutralized. In this case, the adjustments carried out will not be valid for another wire.If the direction of the pendulum always follows the position of the suspension no matter wich one. It can come from the level of the suspension, the wire, the centering of the wire or the fastener of the wire. To know if it is the wire, turn it of a quarter of turn by leaving the fixed suspension. If the pendulum follows the movement, it is that the problem comes from the wire or the fixing of the wire. It may be that the fixing of the wire is offset a little.

To know if it acts of a defect of the suspension, turn only the cable guide. If the pendulum follows the movement, we know the faulty one. In this case, last sapphire is notched or badly crimped. To know if it acts of the fastener of the wire, turn-there while keeping the fixed complete suspension. If nothing occurs, the problem is in bottom. It should be known that an electromagnet whose level is false would be this only of 0.2 millimetres will make station the pendulum. The first pendulum that I manufactured could function only if one slipped a small paper sheet under the coil at a very precise place. This to give an idea of the importance of the levels.

– The pendulum turns in the right direction, makes complete turns on itself, but does so taking longer than it should. It is as if it is lingering on the way. So there is still a small error that counteracts the effect of the earth’s rotation. A pendulum
that acts in this way is very sensitive, and a little can make it work… or stand still. It is therefore necessary to fine-tune the settings. With the habits of the other days, it is now easier to detect disorders.

Example of a pendulum lingering on the way

The pendulum turns in the good direction, makes the full rotations on itself, but takes more time than it should.
Our pendulum finally start to become a real pendulum of Foucault but there is still a small error which contrist the effect due to rotation of the ground (Coriolis force). With its approach, it clock pains to turn freely as if it “walk against the wind”. A pendulum which acts this way is very sensitive, and one nothing will be able to make it function… or station. The adjustments thus should be polished. With the practices taken the other days it became now easier to detect the disorders.

Example of a pendulum “walking against the wind

The pendulum works well, but makes a small ellipse in a certain position, always the same. It can even happen that this ellipse gets bigger and the pendulum stops.
The axis of the electromagnet is somewhat offset. It must be moved perpendicularly to the plane of oscillation of the pendulum. This defect is insidious, because the problem is only visible for half an hour on a complete rotation. And therefore never at the same time..

The pendulum functions well, but makes a small ellipse in a certain position, always the same. It can even happen that this ellipse develops and that the pendulum stops.
The axis of the electromagnet is somewhat shifted. It must be moved perpendicularly compared to the plane of oscillation of the pendulum. This defect is insidious, because the problem is visible during only a very small period in a complete rotation. And thus never at the same hour…

5) The pendulum works perfectly.

Don’t touch anything else. It’s difficult to get it to work properly, but curiously few things disturb a Foucault pendulum once it is well adjusted. You don’t even have to worry about draughts, ground vibrations, magnetism or other phenomena. No precautions are required when throwing the pendulum: aim at the centre and that’s it. All my pendulums work without problems in an open workshop with draughts and people passing by.

Much has been written about the disturbances that affect the proper functioning of an eddy current pendulum. The most common mistake is to believe that the disturbances are external to the pendulum. It is often thought that the earth’s magnetic field, for example, disturbs the Coriolis effect and that ferrous metals should be avoided in construction. I thought so too, until I made a pendulum with an iron balance and it worked perfectly. Never look elsewhere if your pendulum doesn’t work straight away: blame one of the settings instead. And always remember that a short working Foucault pendulum is a kind of miracle.

5) – The pendulum functions perfectly.

Especially do not touch anything more. It is difficult to manage to make it function correctly, but curiously few things disturb a Foucault pendulum once its going. You do not need to fear draughts, vibrations of the ground, magnetism or other phenomena. No precaution is necessary to launch the pendulum: aim at the centre, and that’s all. All my pendulums function without problems in an open workshop with people and draughts who pass.

Many things have been written on the disturbances affecting the performance of Foucault pendulums. The most common error is to believe that the disturbances are external to the pendulum. It is often thought that terrestrial magnetic fields, for example, disturb the Coriolis effect and that it is necessary to avoid ferrous metals at the time of construction. I also believed it, until I manufactured a pendulum whose pendulum was made out of iron and it functioned perfectly. Never seek elsewhere if your pendulum does not function immediately: accuse one of the adjustments instead. And always remember that a short Foucault pendulum that functions is a kind of miracle