I've been working on this project which measures pendulum position and motion. Currently I use a 9-DOF inertial measure unit (IMU), like that used in this project/question. The method works very well, but I'm finding that this approach leads to other complications in my application. That being said, I've been brainstorming alternative methods.

The method/idea that I'm considering would make use of an electro-magnetic field. I'm not an electrical engineer, so maybe what I'm looking for exists, I just don't know what to call it. At the same time, I don't know the technical language/terms so I'm just going to give an elementary description of what I'm envisioning.

What I envision basically extends the concept of this simple circular potentiometer with the ability to infer azimuth and tilt position of the pendulum. OK, bear with me...imagine a doughnut or (bagel if you will) that represents the base of an electro-magnetic field. The pendulum rod would be secured in the middle of the doughnut, yet free to swing. Located on the pendulum rod would be an object which disturbs the electro-magnetic field (a magnet or coil??). When the pendulum is static, the electro-magnetic field is balanced, as the object is located at the concentric center. As the pendulum swings, the object causes the electro-magnetic field to change. Ideally the strength of the electro-magnetic field would be proportional to how close the object was to the doughnut (i.e. give the tilt angle) and the azimuth angle (0-360) could also be inferred.

Based on previous research, I think this would be similar to a Hall sensor? If so, do they make "circular" Hall sensors that would match what I'm envisioning above?

I'm completely open to ideas, comments, and/or suggestions. The main thing I like about the idea above is that I don't have to run a cable carrying data and power to the IMU on the pendulum. Instead, all the electronics are "in the ceiling" and I passively observe the object on the rod.

  • \$\begingroup\$ Are you able to select any convenient location vertically to mount a "ring" within which the tether swings? Do you control the tether itself (can you replace it?) What exactly does the pendulum "bob" look like? Over what distance does it travel? You could also say something about the application, perhaps? Rather than leaving everyone in the dark? \$\endgroup\$
    – jonk
    Commented Jun 13, 2018 at 4:55
  • 1
    \$\begingroup\$ With a magnet on the pendulum, and two orthogonal Hall sensors in the donut, you should be able to get a unique pair or readings for any pendulum position, which you'd then need to turn into an actual pendulum coordinates. Sensitivity would be an issue, the stronger the magnet the better. It would be sensitive to environmental noise, I'd imagine several tonnes of ferrous material moving in the vicinity (a bus using a nearby road) might give a detectable change of reading. \$\endgroup\$
    – Neil_UK
    Commented Jun 13, 2018 at 6:16

2 Answers 2


Yes, they do make such sensors already (google "angle sensor"). Typically, those contain a pair of orthogonal magnetic field sensors (Hall, GMR, you name it) and a small micro-controller which converts the raw measurements to angle using sin/cos transformation. There are cheaper varieties which will only give you raw sin/cos values, as well as more expensive varieties which can handle more DOFs.


Use a web-cam and some computer vision techniques to extract the position versus time of the pendulum. I did it for a small project and it worked flawlessly.

Here is a video of the setup: https://www.youtube.com/watch?v=RyCfrJKnYRM&feature=youtu.be

This shows the position as a function of time:

enter image description here

and this shows the exponential damping of the system

enter image description here

Everything matched physics perfectly.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.