Conception and Design

Final Project



In a seemingly unconnected flow of things, I have another entirely new final project proposal. That's what happens when "anything" is what you have to choose from. I was thinking of creating a hover board that does not depend on having a magnetized floor that it is bound to. My idea is to have the metal platform that the electromagnetically hovering board floats over, be motorized by wheels. As the bottom board moves, so will the top, hovering board. The issue is that the hovering occurs via repulsion. The trick will be how do I get the bottom board to propel the top board along with it. I plan to reach out to my physics professor to come up with ideas. One though is to have repulsion in the middle of the top board but have the outer regions attracted to the bottom board. This would create an arched vector field and may actually bend the top plate, depending on its rigidity.

Electromagnetic Forces

Background & Research

After the outputs week my final project finally took a definite direction. I will be prototyping a hover board by running a current through an electromagnet and using magnetic field sensors to determine whether the current should be decreased or increased to hover the object. Not one of those hoverboards that rolls around on the ground that took over the world in 2015 and 2016, but a real, Back-to-the-Future type hoverboard.

alternate text

By winding coil around steel and running current through it, you can create an electromagenet. The first step was to wind my magnet. The magenet is strengthened by a factor of N, the number of turns around the coil, and weakend by a factor of 1/l, the length of the magnet. Since hand winding is inefficient and would take a long time to get a lot of turns, Rob in the Harvard shop helped me to make a setup where I used a screwdriver, clamp, and electric drill to make a homemade coil winder. The end result was an electromagnet with about 17 Ω resistance.

For my cicuit board, I was able to use Neil's design for the speaker and use it for the electromagnet, simply connecting the magnet to the board instead of the speaker. I milled the board using the Roland CNC then soldered the various parts onto it.

alternate text

For my power source, I added a ribbon header connected to a 9V wall power adapter, for one of the 2x2 headers. It is important to actually connect this properly so I had to test the connector to make sure power and ground were properly alligned.

The next part was to upload a simple code onto my board that would feed current into the magnet for half a second, then stop for half a second. With a 9V source and 17 Ω resistance, this should be about half an amp of current. How we should see the magnet work is that it will attract a piece of metal on and off, alternating every half second. After uploading the code, this would work for approximately 2 minutes before the MOSFET would die. The MOSFET was set up to handle up to 1.7 Amps so it should have been fine with my setup. I replaced the MOSFET 3 times before enlisting the help of electrical engineering experts, Brian Plancher and Mike from the Electrical Engineering lab at Harvard. Brian realized that we needed to add a diode to the setup to offset all of the back emf that resulted from pulsing the current fully on and off. We connected an LED to the setup and what resulted was that the MOSFET would almost immediately start smoking and die. Stumped, Mike came in to help us figure out that the diode actually needed to be connected with the cathode to the anode not anode to anode. Here are the diagrams he drew to help explain the problem.

alternate text

As we approached closer to the deadline for our final projects, I had to modify the project so that the object would be suspended from the top, not levitating from the bottom. There is a lot more domecuntation of work done on attractive levitation rather than repulsive. I'll save the repulsive levitation project for a later date.

Final Project Components:

  • Electromagnet
  • Controlled Current via Microcontroller
  • Hall Sensor Inputs
  • Program to modulate current for levitation
  • Permanent Magnets
  • Housing for all the components

The most difficult part of this project, for me, is figuring out a code that will modulate the current source so that object I have is levitated. This will require inputs from the Hall sensor to determine whether the current should be increased or decreased to keep my object suspended.


Making the Electromagnet

The electromagnet I made for the input devices week was only about 18 Ω. For my final project, I made a new, much stronger magnet. Initially, I designed this spool in Fusion to wind copper around to make my electromagnet

I was planning on 3D printing this and incorporating it for my final design. However, since the 3D printed object would be made of plastic, it would probably lead to a weaker electromagnet. Instead. I used a steel nail, bolts, and washers to create a larger, stronger electromagnet. The resistance of this new magnet is about 50 Ω. This would have taken hours if I had to manually had to wind the electromagnet. Although we did not have an automatic winder in our shop, I was able to use a screwdriver and electric drill to quickly wind my magnet once more. This worked amazingly well and was cool to watch. You just have to be careful with the wire and your fingers since it becomes painful giding the wire up and down the screw. To alleviate this, I used the small red cloth to protect my fingers while I guided the wire.

alternate text
alternate text



Rather than networking my hall sensor board and my elctromagnet board, I created a new board that would both modulate the current through my electromagnet and receive input from the Hall sensor.

alternate text