POV based Magic ball

In this project, the user can ask the ball a yes/no based question and the ball will display the answer in form of a sphere which is achieved using persistence of vision. The sphere display has been created by spinning semi-circle shaped LEDs around a central motor shaft. The rotational speed of the LED’s is fast enough such that the human eye perceives a two dimensional image. The software part of the project basically calculates the refresh rates of the LED so it seems like a 2D figure.


There are 19 LEDs consisting of the semi-circle and the entire display area is mapped to a 19 by 90 matrix where each element in the matrix represents a pixel. The device is powered by a ATMEGA128p connected to Led subsystem and the mic circuit. Whenever the Atmega detects an end of the question, the sphere displays a randomly generated yes/no as an answer to the question. Another feature about our device is that the LED array can generate quite bright light, people can see the display without having to turn off the light. A project which can be made to amuse people in your free time and also to learn the concept of persistence of vision.


Movable Alarm Clock

This project is for the people who faces difficulty in waking up early in the morning or have the habit of snoozing the alarm. In short, this alarm clock runs away from the user as he tries to snooze his alarm. The whole unit is mounted on a chassis with caterpillar-style wheels driven by two small DC motors. The runaway feature is implemented using a proximity sensor which is placed near the snooze button. It has all the features of a “regular” alarm clock: settable time and alarm, snooze, and alarm on/off.

The project is divided into four levels namely the clock feature with display. The alarm feature motion control and proximity detection. The alarm tone is generated using a 555 timer while a real timer kernel is implemented on the Atmega MCU to run each of these tasks independently and simultaneously on the microcontroller. The display is built out of seven segment LED’s running on the principle of persistence of vision. The accuracy of the clock is pretty good, and the speed of the bot can be raised up to 10m/s. The source code of the project is available on the project website along with the necessary libraries.


Laser based projector

This is again a final year project by Melissa and her team, who all are students at Cornell University. The electrical part of the project consists of Galvanometer and the microcontroller which acts as an input to the position detector sensor and also are responsible for control loops which have been implemented using proportionate derivative controller. The mechanical part consists of a combination of different devices which makes up the laser projector. A part of the software runs on a Linux based PC which handles file parsing, point scanning, frame display, and serial transmission to the control board.

For this laser projector, the moving and still parts of the galvo are switched around. The current running through the coils causes the rotary motion of the magnet, and subsequently, both the position detector electrode and the mirror attached to the ends. The project have been designed keeping in mind that further research and development should take place on a similar concept. The project schematics needs a little bit of improvement as accessing the board is a challenging task. A different project but still a long way to go before we see something like this in the market.


Autonomous Goalie

This project will definitely be liked by people who loves playing air hockey. In short term, it’s a low cost robot which detects a puck coming towards it and moves in its required direction to stop it. The great point of the project is, it’s highly responsive and doesn’t use any robotic arms to stop the puck. The heart of the project, an ATMEGA 1284 MCU, takes in echo patterns from each of the sensors, interprets them and accordingly actuates the robots motors. Moreover, it requires negligible human input and can also be powered from a battery based source.

Due to multiple tasks running simultaneously, they have used a tiny real time kernel to execute task in a coordinated manner which has made the robot more responsive to the puck. In the final design, with all the sensors mounted on the robot, a threshold of about 34cm or 1 foot was observed to do the best job of detecting the puck and resulting in proper functioning of the robot. A low cost robot doing a job of the goalie in an very efficient way. However, since the puck is round in shape, there have been certain difficulties in picking it up due to echo not being reflected back.