SIGNAL ACQUISITION SYSTEM DESIGN FOR FAT DETECTION DEVICE IN MILK USING NEAR INFRARED SPECTROSCOPY METHOD

• Main Author: Sholihat Geavonny, Dewanti
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/54331
• Institution: Institut Teknologi Bandung

Milk is a white liquid secreted by the mammary glands in female mammals for food and a source of nutrition for their children. In about one glass of cow's milk (220 ml), there are about 122 calories, 4.8 grams of fat, 115 milligrams of sodium, 12 grams of carbohydrates, 12 grams of sugar, 8 grams of protein, and 293 calcium. A glass of milk contains about 4% fat by weight. In addition, some types of milk are naturally or processed to have a lower fat content, such as nonfat milk (0%) and low fat milk (1-2%). Broadly speaking, more than half of the fat contained in milk is saturated fat. Saturated fat is a type of fat that generally comes from animals, such as poultry, red meat and dairy products. The high intake and levels of saturated fatty acids have implications for an increase in low density lipoprotein (LDL) blood cholesterol. LDL then accumulates on the walls of blood vessels, causing constriction of blood vessels so that saturated fat has a high risk of atherosclerosis and cardiovascular disease. This makes it important to control the intake of fat consumed. One source of fat that is important to control your intake of is milk. In this final project, the design and realization of a measuring device for measuring fat content in milk is carried out. The device is designed to display the voltage from the photodiode that captures the light from the LED that has been passed on to the liquid to be measured. The working principle of this device is that the light from the LED with a wavelength that matches the transmittance spectrum profile of fat passes through a cuvette filled with milk containing fat. Then, the light passing through the cuvette will be captured by the photodiode. Photodiodes that capture light will produce a certain value, depending on the fat content in milk. The value of the photodiode is then converted to voltage and displayed on the OLED display connected to the ADC. The test results show that the displayed voltage value is inversely proportional to the percentage of fat in milk. The higher the fat content, the lower the light intensity that is captured by the photodiode. The low intensity of the light captured by the photodiode causes the resulting voltage to be lower. The results obtained show a non-linear value for certain conditions, so that further predictions using Machine Learning are needed.