Tips for a successfull D.I.Y. Pendulum                                    Latest change 2026-05-12
 
Required skills:

Patience:
Experiments with a Foucault Pendulum generally last days or weeks. The good part of the message is that you can do something else in the mean time, but only if you have the means to precisely log the behaviour of your pendulum. My approach with electronics and software allows for this.

Metal working:
Many parts needed to construct a working pendulum are not available in regular shops. You simply need to make them yourself. I regard (access to) a lathe and a stable column-type drilling machine as a minimum. A milling machine makes several jobs to a pleasure.

Wood working:
People who are used to d.i.y. work in the home often have enough tools and experience for the woodworking required. Manufacturing the coil members to the required precision may ask for some improvisation.
Some parts may lend themselves to maker processes like 3D printing, laser cutting etc.

Electronics:
You don't need to be a professional for this, but basic understanding of working with operational amplifiers, Ohms Law, Voltage and Current levels, etc. is required. Basic experience with (and access to) an oscilloscope is as good as mandatory.

Software:
It is almost inevitable to (be able to) modify the Arduino firmware and the PC software I present here to your particular needs. Or completely design your own. So if you don't know already, learn programming, or find a friend who knows to. I use (Arduino) C, C++ and Free Pascal.

Study the theory as good as you can. I am certainly not a skilled mathematician or physicist, I am a retired electronic engineer and yes that means a certain math background, but most of the math I've seen in the articles I mention go beyond my skills. Not withstanding that I could mostly "read between the formulas" and understand the messages there.

Suppress (the effects of) the elliptical path any pendulum tends to follow.
I use only the method according to the Schumacher article with rather good results. That means that the timing of the drive-pulse is extremely important.
I have no Charron ring or some similar mechanism to reduce the amount of ellipse. My experience so far is that a correct timing also seems to limit the amount of ellipse growth.

Use a drive coil of small diameter. The Schumacher equation predicts that a bob with a high Q needs to be driven very close to the center. So, to make your coil effective, give it around the same radius. However, you generally won't know the Q in advance and maybe need to redesign your coils after some experiments.

Do not use an iron core in your drive coil or in any other coil in your system, at least not if you have a magnet in the bob. The amount of energy to pump into the system each halfswing is generally so little that it can be done with a coil of only copper wire. The presence of iron in the center of the floor unit can easily cause big problems..

Use a Center-Detect coil of sufficient diameter such that it will give enough signal, even when a substantial ellipse has developed.

Do implement a Rim Coil as I did. Give it a radius of around 2/3 of the intended amplitude of your pendulum. It will be very helpfull in accurately determining and controlling the amplitude of the pendulum. The Schumacher method requires a stabilized amplitude.
It must be said: in the literature I've never seen an explicit mechanism for active amplitude control, and I do not know how such a mechanism could influence the performance in other ways. I've planned experiments to find out.

Do implement a position measuring system. I regard it as the only way to accurately monitor the behaviour of your pendulum. Currently I use the system with Hall sensors and a magnet around the wire. Earlier I mentioned a system based on cheap optical sensors, but that did not work well. The sensors, although with analog output had a step-wise tranfer function, to rough for the purpose.
I have now (july 2025) a system based on AC capacitive coupling developed for my sub-meter pendulum. It is available as a kit with electronic circuitry.
An article about this PMS is expected to appear in the electronics magazine Elektor by mid 2026.

Use a simple clamp to fix the wire at the top. There exists a variety of bearing constructions with balls, cylinders, knife-edges etc. These are all difficult to construct with the required precision and / or may introduce anisotropy or friction which -even in the slightest amount- causes problems like developing a substantial elliptical path or complete absence of Foucault Precession because of stalling at some angle.  

Extremely important is the rigidity of the top mount structure. The slightest wobbling of the top mount causes ellipse growth or even stalling of the precession at a certain angle.

Fix the wire such that the weight of the bob is taken by some construction above the clamp (e.g. a winding axle). The clamp then only cares for the horizontal fixation of the wire. This way there is no need to tighten the clamp very strong with the risk of cutting into the wire.
This also eases the height adjustment of the bob.

I use Piano Wire, also known as Music Wire and family of Spring Steel. Piano Wire has a tensile strength around 6 times that of normal iron wire and is manufactured with very narrow tolerances. It is made to withstand all tortures of master and amateur pianists without going out of tune.
It may be desirable to use a piece of wire which has never been rolled up (or does not remember that) for the upper part of the wire. I have indicatons that a wire which takes a curvature when left untensioned can contribute to ellipse growth.
I once used normal iron wire of 1mm diameter (tensile strength 36 kg/mm2) with a bob of 6 kg and a period of 4.2 seconds. It broke after around 12 days of operation because of material fatigue.
Stranded steel cable can be a good alternative, but is probably not easy to aquire. A disadvantage of stranded cable is that it lengthens and rotates when it is loaded, and that it has a low torsion spring constant. This may cause your bob to rotate along its axis for long time.

Try to use a bob as compact as possible. The weight itself is not that important, although mass goes with the 3rd power of linear dimension and surface (as seen from the side and determining air friction) goes with the 2nd power of linear dimension.
A bob made from high density material is smaller and has less air friction compared with a lower density bob. I use lead, which is quite dense (11 kg/dm3) and modestly cheap. Besides that you can easily melt it and pour it into your bob's shell.
Also the shape of the bob is not critical. A sphere might be the optimal compromise between mass and air friction, but is hard to manufacture. A cylinder is much more easy to fabricate. My pendulum appears to have a Q over 3000.

Log as many parameters of your system as you can think of. In case of problems cross correlation of otherwise redundant data can tell you a lot about the (dis)functionng of your system. Take care that you can analyse the data in an easy and flexible way. I wrote special software for that.

Put a wind screen around your pendulum. My experiments have shown that the pendulum is very sensitive to even imperceptable air draught from e.g. an airco or other mechanical ventilation systems or house heating systems. Use a candle or so and check at the location of your pendulum that it has a steady flame. If the flame flickers or wavers your pendulum may be in trouble.

Do not believe everything I write on this site. Sometimes I am wrong.