Inverted Pendulum Project

I built an inverted pendulum for the final project in my Advanced Digital Design class. I had the cart mechanism and motor. I just needed to build the circuit to drive the motor and interface it to the FPGA.

Here's the Powersupply and control circuit. The circuit on the breadboard wasn't very neat, but it worked. I wanted to redo it but it probably wouldnt have worked afterwards.

This is the final schematic of the Hbridge. The circuit had built in current limiting as well as limit switch inputs to shut down the motor if it got out of control.

  I found out the hard way that the circuit needed current limiting. Some mosfets went bad which shorted out the powersupply, almost catching the breadboard on fire.

  A rotary encoder was used on the pendulum and motor to provide feedback to the control system.

(Here's a link if the embedded video fails to load not_working.mpg)

Here's a video from the initial testing. After some quick troubleshooting, turns out it was a sign issue. The motor would move in the wrong direction to correct for the pendulum falling.

(Here's a link if the embedded video fails to load working006.mpg)

Here's a video of it working. It was still pretty jittery, and unfortunitly this was the best we could do. The original plan was to get all of the control logic built into the FPGA. Due to time constraints and the difficulties involved in building the driver circuit we had to simplify the project.

We used the FPGA to count pulses from the encoders and generate the PWM signal to the motor driver. The FPGA would send the count values to a PC using RS232. The PC would pass these values into Matlab which would generate the new value for the motor.