HARDWARE & MECHATRONICSIN PROGRESS

Maglev Linear Motion Platform

An experimental magnetic levitation platform with analog control circuits, Hall-effect feedback and hardware designed in SOLIDWORKS. I'm building it with Prof. Kwee Yan Teh at SJTU.

ROLEStudent researcher (IPP)

TEAMIndividual project with Prof. Kwee Yan Teh

TIMELINEFeb 2025 - Present

CATEGORYHardware & Mechatronics

1D vertical

STABILIZED AXES

Hall effect

FEEDBACK SENSOR

Analog + Arduino PWM

CONTROLLER

Overview

Magnetic levitation looks like magic, but it's really a control problem. An electromagnet pulling a permanent magnet upward is an unstable system. If the magnet gets too close it snaps to the coil, and if it drifts too far away it falls. To hold the gap you have to keep measuring the position and correct the coil current faster than the error grows.

I'm building that control loop from scratch as an Individual Project Pursuit (IPP) with Prof. Kwee Yan Teh at SJTU. There's no maglev kit and no off-the-shelf controller involved.

What I built

Analog circuits for 1D stabilization. A Hall-effect sensor measures field strength, which tells you roughly how big the air gap is. That signal gets conditioned and used to drive PWM-controlled electromagnets with proportional control.
An Arduino version of the MIC502 PWM controller chip. The original IC worked, but changing the tuning meant swapping resistors and capacitors. With the same behaviour in firmware I can change the loop parameters in code, which is a lot faster.
Several electromagnet mounting platforms and supports in SOLIDWORKS. Most of the design changes were about keeping the solenoid axis vertical, because any misalignment shows up as sideways drift that a 1D controller can't correct.

Why I like this project

It needs electromagnetics, analog electronics, embedded code and mechanical design to all work together. If any one of them is off, the platform falls, so you find out very quickly when you've got something wrong.

Current state

The platform levitates stably in 1D with the Arduino controller. Next I want to measure how well the loop handles disturbances, and then work toward linear motion along a rail of coils.

APPENDIX A // MEDIA

FIG.01 · Acrylic test frame with the electromagnet mount — FIG.01 · ACRYLIC TEST FRAME WITH THE ELECTROMAGNET MOUNT

FIG.02 · Control circuit: Hall sensor to Arduino to H-bridge to electromagnet — FIG.02 · CONTROL CIRCUIT: HALL SENSOR TO ARDUINO TO H-BRIDGE TO ELECTROMAGNET

FIG.03 · Test bench with the breadboarded controller and bench supplies — FIG.03 · TEST BENCH WITH THE BREADBOARDED CONTROLLER AND BENCH SUPPLIES

FIG.04 · PWM output from the Arduino at about 50% duty cycle — FIG.04 · PWM OUTPUT FROM THE ARDUINO AT ABOUT 50% DUTY CYCLE

FIG.05 · Hall sensor output with swift rise time and fall time at 50% duty cycle — FIG.05 · HALL SENSOR OUTPUT WITH SWIFT RISE TIME AND FALL TIME AT 50% DUTY CYCLE

FIG.06 · Chasing switching noise spikes on both channels — FIG.06 · CHASING SWITCHING NOISE SPIKES ON BOTH CHANNELS

VIDEO SLOT · AWAITING UPLOAD

FIG.07 · LEVITATION DEMO, VERTICAL STABILIZATION