Nuts and Volts - May/June 2018

This time, we’ll expand the LED matrix to run the

182 LEDs that the clock uses and make a printed circuit board (PCB) for the clock. Then, we’ll wrap it up with a discussion of how the software works and put the finishing touches on our unique timepiece. Let’s get started.

The Printed Circuit Board

I decided a PCB would be best to have for the final circuit, due to the large number of LEDs and the need to arrange them and the resistors and diodes in a circular pattern. I wanted to manually route the traces so that the only traces visible from the front of the clock are the Charlieplexed lines, which would be laid out as concentric traces on the front of the board among the LEDs. I played around with a number of PCB design programs and decided to use Dip Trace as I found that it could most easily do what I wanted.

Dip Trace has a schematic capture module where I created the schematic for the clock. Then, Dip Trace transfers the schematic to the PCB layout module. To arrange the LEDs and other components in a precise circular pattern, I created a spreadsheet in Excel to calculate the position and rotation of the components. I entered these values into the Properties field of each component. Once all the components were in place, I manually routed the traces. Dip Trace has a 3D rendering tool to show what the completed board will look like (Figure 1).

I used Bay Area Circuits (see Resources) to make the circuit board as they had a student special that allowed me to make a single board for a very reasonable price. Dip Trace generated the necessary files to send to Bay Area Circuits so they could make the board. While I was waiting for the board to be made, I ordered more parts (mostly LEDs) to build the clock.

Building and Testing the Clock

It took me about four hours to solder all the components onto the PCB (Figure 2). Once I was done, I adjusted the code that I had written for the test circuit to work with the additional LEDs and button. It turned out there were about a dozen or so LEDs that wouldn’t light up. I took note of which ones were not coming on and realized their anodes were all connected to the same line.

I checked that line for continuity to the current-limiting resistor and to its I/O pin on the microcontroller. That all looked good. Then, I thought I had a bad microcontroller, so I replaced it. I was pretty confused when that didn’t fix it either. I checked the code, but couldn’t find any issue there. Finally, I realized what the problem was.