MAGNET LEVITATION EXPERIMENT

**mariost** · Apr 22, 2019, 05:55 AM

I use a coil (450 r-0.5 mm) a hall sensor inside the coil and a mosfet as a switch for the current to the coil. The diode protects the mosfet from the back emf of the coil. A resistance controls the current to the base of the mosfet. The coil attracts the magnet , then when the supply stops the magnet begin to fall , the supply to the coil begin again and attracts the magnet again .This happens many times per second and the magnet seems to stay in the same place.
I try to find a shematic for the opposite effect. The object that levitates to be up and the electromagnet to be down.I' ve seen a system of 4 coils and 2 hall sensor but I cannot find it's shematic .Doew anyone can help? Thank you. You can watch this experiment here:

**Scotsman Hosie** · Apr 23, 2019, 02:28 AM

Ingenious use of the Hall sensor. Kind of the "Segway" of magnetic levitation, eh?

**JTG** · Apr 27, 2019, 02:06 AM

K and J Magnetics had a blog post a while back where they built the inverted setup with ratiometric sensors, but they didn't provide a detailed schematic. I get quite a few hits when searching for "repulsive electromagnetic levitation schematic," but there are about as many ways to go about doing it as there are people building them.

**mariost** · Apr 27, 2019, 08:39 AM

I 've found two shematics but there are a few correctionts from other vewers so i am not so sure that they are right.

**Scotsman Hosie** · Apr 27, 2019, 04:01 PM

Originally Posted by mariost

I 've found two shematics but there are a few correctionts from other vewers so i am not so sure that they are right.

I wonder if the circuitry commonly used for switching power supplies couldn't be used – or adapted – to work with that set up? (Just a thought.)

**jdurand** · Jun 3, 2019, 07:03 PM

You could easily feed the sensor into a small microcontroller like an LPC80x series or even an STM8 series. That would smooth out the detections and output a PWM signal to drive the coil. You'd have at least part of a PID loop in the processor.

You could also sense the location of a shiny object like a ball bearing with an infra-red photo beam. That would give you an analog position of the ball and make smooth control easier.