People take clocks for granted. In fact, no one really looks at them, a clock simply displays time. I doubt many people have wondered about the inner workings of a digital clock, it is just a bunch of wires, right? I decided to make one because I needed a simple, easy to complete project.

    There are several types of clocks and multiple ways to build each type. Some clocks keep time by tracking a swinging pendulum or by using clockwork. Others keep track of time with specific electrical circuits. Electrical clocks then have two subgroups, micro-controller based and a cheaper type normally found on store shelves. I weighed my options and chose to build the simpler but more expensive kind. The electronics I put together just displayed the time. The micro-controller, basically a simple computer, kept track of time.

    At the time I had built a few quick circuits involving simple displays, such as a single LED or a seven-segment display, a set of seven LEDs meant to display a number. I wanted to use a more complicated display, one with more lights than output pins on my micro-controller. A programmer can control an output pin (an electronic switch) with code. They can decide when it will send power and when it will not. The micro-controller I used has only twelve output pins, which causes an interesting problem: how do I independently control thirty lights with only twelve switches?

    An integrated circuit, or computer chip, can provide the answer. After searching the Internet for awhile, I finally remembered about the shift register. I first learned about this chip in a tutorial about making a bar graph out of LEDs. Shift registers, or small integrated circuits designed to take pulses of electricity and then split up the stream and send each value to a separate pin, effectively allow me to use one output pin on my micro-controller to control the eight output pins on the shift-register. At this point I had enough output pins to properly control my display. However, the problems do not end there.

    The makers of the display did not want to waste space by providing upwards of sixty input pins (each light needs two pins, one for positive one for negative) so I could control each light individually. Instead the display has only twelve pins broken down as follows: seven pins, one for each segment, four more pins for choosing which number to light up, and one negative pin shared between all of the lights. The segment pins, when powered, turn on a specific light. These lights are seven bars arranged in the shape of the number eight, just like a commercial digital clock. When I power the pins in the correct configuration, I can display a limited set of letters or any number. Together these seven lights make a digit; there are four of them on my display. The four digit choosing pins then decide which of the digits to turn on. All of the lights channel the electricity back through the single ground pin.

    Because the display only has one set of segment pins, only one digit may be displayed at a time. For example, if I were to send data to the display that would create the number two and I powered all four of the digit choosing pins, all of the digits would display the ‘two’ character. This happens because the digits share segment, or data, pins. They cannot all display a different character at the same time. This design saves space so nearly every type of display uses it. The ubiquity of this problem made the solution easy to find.

    The answer, called multiplexing, involves consecutively flashing each digit on and off then changing the display data before lighting up the next digit. A computer can perform this task so quickly that all of the digits appear lit at the same time; they can even have different numbers on them. Theoretically, this works; however, real life electronic components require a specific amount of power to work as intended.

    The micro-controller, to accommodate a wide range of devices, puts out more power than my display could receive. So I needed another part to lower the power. A resistor can make this happen. I can only explain resistors by using water as a metaphor for electricity. I like to think of them as a thin section of pipe. So when the water gets to the thin pipe, the water slows down just like the current of the electricity. Each light pin needs a resistor between it and the shift register so it stays cool and does not prematurely break. With the hardware completed, only the software remains.

    The micro-controller requires some code to complete tasks. Basically it needs to keep track of time, then from that value determine each digit to display and finally invoke the multiplexing strategy to display the numbers. It only needs a programmer to provide the necessary code.

    Buttons to set the time make a useful addition to any clock. I imagine the insides of a button looks like a removable section of wire. When pressed, the wire drops down and a connection occurs; when not pressed, the wire lifts up with the button and the connection stops. In reality buttons lift and place a metal ball onto a set of four pins. To use a button, I only had to provide power to the input pin then connect the button’s output pin to an input pin on my micro-controller. The micro-controller can then check the button’s state by checking if the input pin has power. Next I added code to increment the stored time value when the micro-controller detects that the button is down.

    I finished the clock so I decided to plug it in and set the time. There it lived on my desk for a few days, until I got tired of the cluster of wires and random components. After clearing the table, I looked at the clock and decided I do not need it. I pushed it into a box and moved on to the next project.