Levitating a ball

[shivkrish22]

hi
i have been working on this project to try and levitate a steel ball(about 30 gms) using the info on this site>>
**broken link removed** as reference.
however i have not managed to get anything working as it should. is there any different kind of controller that i can use?? (probably using microcontrollers or thru the PC). has anyone tried it out??
The exact schematics will be very helpful as i have to complete it by the end of March.

 

[tansis]

A micro-controller is not a magic cure all shivkrish.


How is your setup "not working as it should" be specific,
is the coil not providing a strong enough field or are the
optical sensors not detecting the ball?

Have you followed Barrys circuit diagrams and instructions?
If you have designed your own control circuit please show us
so we can check it for errors. Also a detailed drawing
showing the position of the optical sensor, better still a photograph
of the project.

 

[Oznog]

The circuit is in itself quite unstable. The key is getting just the right response taking into account the speed, position, and inertia of the object versus the strength and time lag of the coil. They describe Bode plots there. What I'm getting at is that a DSP microcontroller can do a bangup job of this, better than analog, IF you know what you're doing. A non-DSP chip may not be able to do calculations fast enough to keep it stable. Well, it's not a real fast device so it may be possible... who knows.

 

[bmcculla]

I did this in one of my classes. We used a Lead compensator circuit that was about 3 opamps.

https://inst.eecs.berkeley.edu/~ee128/fa04/labs.htm has the magnetic levitation labs. Lab 2 shows the lead compensator circuit - you will have to calculate values for your system though.