Some years ago, Steve saw an 8x8x8 led cube capable of displaying 3 dimensional patterns and images. The cube was controlled with several PIC processors and supporting hardware. This project will be expensive; the cost per tri-color LED is over $1 and there are 512 LEDs per full cube. The original designers used High-end PICs to drive the cube, but we can do better using an FPGA to drive the system along with concepts we have learned in our engineering coursework.
The system consists of common-cathode LEDs, and a driver to pulse the LEDs at a high frequency. However, with the enormous number of diodes combined with the color depth required to produce decent images, may be to much for a PIC. An FPGA will work in parallel, only needing a NIOS soft processor or possibly a simpler program to provide user input. Perhaps this system could also be combined with some sort of audio signal interpreter (as seen in MS-media player 9+)?
At 50.0 MHz, the NIOS II alone could provide 1,100 ASM instructions (color levels) per led, per color channel, per frame of animation (512 LEDs, 30 frames per second, 3 colors per led). Combine this with a hardware driver to take most of the work from the processor, and a lot is left open to the user.
Our idea is based off of what Steve saw before. We want to make a 3D Hypnocube prototype that displays interesting animations. We looked at some online demos of previously built Hypnocube projects, and believe it would be an interesting and challenging project. Options we are considering include making the cube rotate to produce images, or possibly creating a 3d pong game with controls so two or four players could play. The pong idea would be very difficult to implement, but we believe that it could be possible depending on how the production of the cube itself goes.
Problems that we foresee are power problems in the system. We don’t think the DE2 board can power the number of lights we plan on using. Also, the lights will be expensive, therefore we have to decide between using RGB lights and possibly having to reduce the cube size from an 8x8x8, or we could use single or dual color LEDs to reduce costs. Other challenges will be the amount of fabrication necessary to implement the cube. The only other issue is the number of chips that will need to be interfaced, especially the master controllers communication with the PIC processors.
There are many features we could add to the cube after we finish the basic control. For example, we could make create animations such as snakes or cascades, maybe some 3D effects if we can get all of the software synchronized..