VCE Systems Engineering study materials
VCE Systems Engineering study resources

Diagnosing Intermittent Faults in a Traffic-Light Prototype

VCE Systems Engineering · Put your knowledge to work in context.

All case studies

Diagnosing Intermittent Faults in a Traffic-Light Prototype

A student has built a working model of a pedestrian crossing light using a 5 V logic board, three LEDs (red, amber, green) each with a 220 Ω series resistor, a pushbutton, and a relay that switches a 24 V lamp. During testing the green LED sometimes stays on after the sequence should have ended, and the relay occasionally chatters when the button is pressed. The circuit is built on a breadboard with long jumper leads. The student's first instinct is to replace the logic board. The teacher asks for a structured diagnosis before any part is swapped, noting that the fault is intermittent and that the 24 V lamp circuit shares a common ground with the logic board.

1. Propose a structured fault-finding sequence, naming the measurements and the order in which to take them.

Read the worked answer

Principle Change one variable at a time; observe before substituting parts.

Step 1 - Reproduce and characterise Log when the fault occurs: after how many cycles, at what button press rate, and whether it correlates with the relay firing. Intermittent faults need a repeatable trigger before they can be isolated.

Step 2 - Check the supply Measure the 5 V rail with a meter (and oscilloscope if available) during a button press. A dip below the logic threshold explains both a stuck output and relay chatter. Also check the 24 V rail under load.

Step 3 - Check the ground path With a shared ground and a relay switching 24 V, ground-bounce is a prime suspect. Measure voltage between the logic board ground pin and the supply ground while the relay operates; anything above a few tens of millivolts is significant.

Step 4 - Check the input The pushbutton is likely to bounce. Observe the logic input during a press; if it chatters, add hardware debounce (RC) or software debounce.

Step 5 - Check the output and wiring Wiggle-test the long jumper leads and reseat them. Check the LED resistor value and the relay coil suppression diode - a missing flyback diode causes switching transients that can reset or latch logic.

Step 6 - Only then substitute Swapping the board first destroys the evidence and may mask a wiring fault that will return.

2. Given the symptoms, rank the three most probable causes and justify each ranking.

Read the worked answer

Most probable - ground bounce / shared ground return The relay coil and the 24 V lamp return current through a shared ground. Long breadboard leads add inductance and resistance. When the relay switches, the ground reference at the logic board shifts, which can latch an output (green stuck on) and cause the relay to drop out and re-energise (chatter). This single cause explains both symptoms, which makes it the strongest hypothesis.

Second - missing or ineffective flyback diode across the relay coil An unsuppressed coil generates a large inductive spike on switch-off. That spike can corrupt logic and produce erratic relay behaviour. It is easy to check and easy to fix, and it is a classic student omission.

Third - switch bounce on the pushbutton Bounce reliably explains relay chatter on press, but it does not explain a green LED staying on after the sequence ends. So it is a partial explanation only.

Ranked lowest - failed logic board The board runs the sequence correctly most of the time, which is not the signature of a dead board. Intermittent, load-correlated faults point to wiring and transients, not silicon.

Judgement Fix the ground return and add coil suppression first; then debounce. Replacing the board first is the least defensible action because it is the least likely cause and the most expensive to test.