Developing a low-cost turbidity sensor using synchronous detection
Author:
Tsugunobu Miyake ’25Co-Authors:
Faculty Mentor(s):
Alan Marchiori, Computer ScienceFunding Source:
Kalman Fund for Undergraduate Research in the SciencesAbstract
Turbidity is an essential measurement to assess water quality. Turbidity optically quantifies the cloudiness of a liquid and is commonly measured in Formazin Nephelometric Units (FNU). Commercial turbidity sensors are expensive, and previous work has developed numerous low-cost turbidity sensor designs. However, these designs face numerous challenges, causing inaccuracy in measurement, such as ambient light and bubbles in the liquid. This research developed an improved low-cost turbidity sensor that lessens the error caused by these challenges by adapting a colorimeter evaluation board, equipped with an analog-to-digital converter that reads the raw voltage output from a photodiode. The lock-in amplifier minimizes the effect of ambient light on the turbidity measurement by differentiating the light emitted from the LED and any ambient light. In lab tests, the developed continuous turbidity sensor was within 0.4 FNU from the actual turbidity more than 80% of the time, and nearly all of the readings fell within 1 FNU from the actual turbidity. Furthermore, analysis of the collected data revealed that the model can be accurately calibrated from only two distant turbidity measurements.