Check: Match the sensor to the quantity

Knowledge check · self-test
Check: Match the sensor to the quantity
A photodiode produces a light-dependent current, which a circuit converts to a usable voltage. Start by measuring empty space and then a known object. Change the distance, angle, surface, and room light one at a time. Choose a threshold only after those readings are visible. Then test a dark object, direct sunlight, and a value close to the threshold. A single clean desk test cannot prove reliable detection. The panels show why a useful threshold must hold across changes in the scene, not just one test. Use a small reading table. Begin with no object. Add one pale object. Add one dark object. Try three distances. Turn room lights on and off. Face the sensor toward and away from a window. Save the raw number each time. Look for overlap between present and absent cases. Set a decision band, not a perfect point, when readings move around. Now repeat the test after the sensor warms up. Repeat it with another unit. Record any missed or false detection. The threshold should follow the evidence and the allowed cost of each error. The practical pattern is to calibrate with the same geometry the device will use. For a bin-fill sensor, mount the ultrasonic transducer at the final lid height and measure empty, half-full, and full echoes. For an IR proximity sensor, test the real target materials rather than only a white card. For a daylight sensor, collect readings at dawn, midday, and night in the enclosure, then choose a threshold with hysteresis so the output does not chatter at the boundary.
A photodiode produces a light-dependent current, which a circuit converts to a usable voltage. Start by measuring empty space and then a known object. Change the distance, angle, surface, and room light one at a time. Choose a threshold only after those readings are visible. Then test a dark object, direct sunlight, and a value close to the threshold. A single clean desk test cannot prove reliable detection. The panels show why a useful threshold must hold across changes in the scene, not just one test. Use a small reading table. Begin with no object. Add one pale object. Add one dark object. Try three distances. Turn room lights on and off. Face the sensor toward and away from a window. Save the raw number each time. Look for overlap between present and absent cases. Set a decision band, not a perfect point, when readings move around. Now repeat the test after the sensor warms up. Repeat it with another unit. Record any missed or false detection. The threshold should follow the evidence and the allowed cost of each error. The practical pattern is to calibrate with the same geometry the device will use. For a bin-fill sensor, mount the ultrasonic transducer at the final lid height and measure empty, half-full, and full echoes. For an IR proximity sensor, test the real target materials rather than only a white card. For a daylight sensor, collect readings at dawn, midday, and night in the enclosure, then choose a threshold with hysteresis so the output does not chatter at the boundary.
Back to Light and Proximity Readout Physics: Practitioner: Transimpedance and ToF Readout