Line Follower Spins Out On Glossy Tiles
Team Bravo's line-following robot uses two TCRT5000 infrared reflectance sensors, an L298N motor driver and four AA NiMH cells giving 4.8 V nominal. On the matte white classroom floor with black tape it follows the line reliably. At the district arena the floor is glossy white tile and the line is shiny black electrical tape. The robot now spins in place at junctions and sometimes stops for a second before lurching forward. The teacher notices the L298N heat sink is hot after a few minutes and that the battery pack reads 4.1 V after five minutes of running, while the motors are rated 6 V. One member suggests adding a fifth cell and raising the sensor threshold in code.
1. Distinguish the sensor problem from the power problem and describe a test that would separate them.
Read the worked answer
Treat them as two hypotheses and test them apart.
Sensor test. Hold the robot up and point the sensor at matte white paper, glossy white tile, black tape and shiny black tape, reading raw values or watching the onboard indicator. If the white tile and the black tape give similar readings, the sensor is seeing a reflection rather than a colour: TCRT5000-type sensors depend on diffuse reflection, and a glossy surface acts partly like a mirror, so the reading changes with angle rather than with the line. Repeating at several heights and tilt angles shows whether the gap between 'line' and 'floor' has collapsed.
Power test. Put the robot on blocks, run the motors, and measure the pack voltage at the driver terminals and at the logic supply. A pack reading 4.1 V off-load may sag far lower during start-up, and the L298N drops roughly 1.5 to 2 V, so 6 V motors see much less. Heat on the heat sink after minutes indicates current and losses, not simply a warm room.
2. Evaluate the two suggestions: adding a fifth cell and raising the sensor threshold.
Read the worked answer
Adding a fifth cell raises nominal voltage to 6.0 V, about 7 V fresh off charge. That gives more torque, but it does not fix a sensor that cannot distinguish line from floor. It also pushes the L298N and the motors harder, so heat and voltage sag under load get worse, and stall current may exceed what AA cells and the driver can safely supply.
Raising the threshold is a partial fix at best. If the glossy surface shifts readings erratically with angle, a single threshold will be correct at some positions and wrong at others, so the robot still loses the line at junctions. It may also create false 'line' detections on bright tiles.
Better: calibrate on the actual arena surface, set a band with hysteresis rather than one value, add a third sensor so the robot can tell 'line to the left' from 'line to the right', and shade the sensors from overhead light. Fix sensing first at 4.8 V, then consider a separate higher-voltage motor supply only if torque is genuinely short.
3. List the electrical and safety changes the team should make before the next run, with a reason for each.
Read the worked answer
- Fuse the battery lead. NiMH cells can deliver large current into a short circuit; a fuse limits the damage.
- Separate motor and logic supplies, or at least use a star ground. Motor inrush current causes the dips that make the robot pause and lurch.
- Add a bulk capacitor across the driver supply. It absorbs start-up dips instead of letting the logic brown out.
- Fit a flyback diode wherever a coil is switched. Without it, switching spikes stress the driver and can reset the microcontroller.
- Respect the motor rating. Do not exceed 6 V nominal on 6 V motors; check the driver's current rating against stall current and give the heat sink airflow.
- Test on blocks, then on a short matte strip, before the full arena. Log sensor values on both surfaces so the next change is evidence-based.
- Secure the pack and insulate exposed terminals. A loose pack that shorts on the chassis is the main fire risk in a school setting.