LED Cube Powered by Microcontroller

1
LED Cube Powered by Microcontroller
Jeremy Kornwitz '09, Tim Beck '10, Brian Koopman
'12, Max Macrae '12 (Sponsor Jerry Breecher)?
Multiplexing The main idea behind powering and
regulating the cube is multiplexing. The cube
has four layers, each made up of 16 lights.
There are 16 vertical columns composed of four
lights each. The columns are hooked into the
high voltage, while the layers are hooked into
the low voltage. A light will only light up if
its column and its layer have a voltage across
it. (So if one layer and two columns are getting
power, only two lights in that layer will light
up)?. By controlling which layer and which
column is getting power, we can turn specific
lights on and off. The virtue of multiplexing is
that it cuts down on the number of wires needed,
but it requires that each layer be activated
sequentially. This avoids unintentional LEDs
being lit.
Construction of the Cube The cube was built in
layers with the help of a homemade jig. First,
each horizontal layer was constructed in a 4x4
grid, connecting the cathodes to create a common
layer. Then, the layers were stacked on top of
one another, connecting the anodes. We ran into
a problem here and had to use 16 individual
connecting wires to connect the anodes. Once the
layers were soldered together the cube was
attached to the perf-board. Finally, four
additional wires were connected to each
horizontal layer so they could be activated.
Coding The code that drives the cube is written
in C, compiled to PIC assembly, and loaded on the
microcontroller. The code uses two sections, the
main level and the interrupt level, to send
voltages to the proper pins to display a
pre-programmed pattern. The two levels alternate
thousands of times per second to create the image
on the cube.
What We Did We constructed a 4x4x4 LED cube
controlled by a PIC microcontroller with the
ability for each LED to be addressed
individually. Using an Interrupt Service Routine
to illuminate LEDs at appropriate times we then
applied our self-designed algorithms to display
unique patterns on the cube.
What is an LED? LED stands for light emitting
diode. The short lead is the cathode, or
negative side. The long one is the anode, or
positive side. LEDs are commonly used in lights,
including household flashlights.
Main Level The main level initializes the
variables to set up an interrupt timer and
contains the information about the patterns.
This code repeatedly sets or clears bits in a
three dimensional array representing all the LEDs
in the cube. If a LED's virtual representation
is set to 1, it is turned on and vice versa. The
pattern being used determines the order the LEDs
are turned on and off.
View of a Layer From the Top
View of a Column From the Side
Cathode
Cathode ends, low voltage.
LED
What is a Microcontroller? A microcontroller is a
small computer with a relatively simple CPU. It
generally consists of a crystal oscillator to
keep track of time, an interrupt mechanism to
provide real time response to the events it
controls, and a small amount of memory.
Microcontrollers are used in virtually all
automatically controlled devices.
Interrupt Level A timer that runs while the main
level is executing causes an interrupt thousands
of times a second, bringing the program to the
interrupt level. Each time the program gets to
the interrupt level, it takes the contents of one
level of the array and turns on the proper LEDs
for that layer. The interrupts happen fast enough
that all four levels appear to be on
simultaneously.
Anode end, high voltage.
Building a Row
The Jig
Transistors (Pictured Below)? The transistor is
the main component that makes multiplexing
feasible. Transistors are designed to manage
electrical outputs in this case the transistor
acts as a gate to a low voltage area. Each of
the four layers is connected to the (1) leg of
the transistor, but no electricity will pass
through the (3) ground leg until the (2) gate leg
is also receiving power. Since each gate leg is
connected to a pin on the chip, specific layers
can be connected to ground at will by the chip.
Hardware The Board In order to run the cube, we
need a power source and something to control the
lights. The chip controls the lights by
controlling the electrical output. In order to
run the cube with minimum power, transistors are
also used. The main idea is that once the chip
has power running across it, the proper program
can redirect that power through the lights using
a set of transistors as the low voltage and the
row of resistors in the bottom right as high
voltage. Pin diagrams tell how to maintain a
voltage across a chip so that it can do work, as
well as naming each specific pin and detailing
their uses. The board schematic also has specific
instructions for connecting the chip to its
programmer. These all come from the manufacturer.
3. Connects to a low voltage (ground).
1. Connects to a high voltage (source).
Wire Diagram
2. Current can only flow from 1 to 3 if this is
also receiving an output (a high voltage from
source).
Cube Attached to Perf-Board
Schematic Diagram
Pin Diagram
The Board