Design Of Light Sensor Using Algorithm Matching To Human Eyes

Light sensor is a device to convert light intensity information to output digital signal, mainly for LCD auto-control. Light sensor has a broad market. It can be used in applications such as consumer electronics (smart phone, PDA, MP3), TV, medicine industry and automobile.Based on the application requirements of light sensor and the requirement to LCD screen luminance of human eyes, this thesis analyzes the shortcomings of the traditional method, which directly converts light signal to electricity signal. Then a new algorithm can be used for the realization of light sensor is proposed. Light sensor realized with this algorithm can realize different conversion resolutions, according to the ambient light. This characteristics match to the response to light of human eyes. At first, this algorithm is verified in MATLAB. Then to implement the algorithm on circuit level, light sensor can realize at most 16bit resolution is designed. The designed light sensor can probe the light intensity range from 0.1lux to10000lux. When at dim conditions (0.1lux~100lux), the resolution of the output signal can reach 16 bit. While for bright conditions (5000~10000lux), it can reach 9 bit. Besides, the designed light sensor can realize the functions of infrared light correction and temperature compensation, which can effectively inhibit the affection of infrared light and dark current to the output result of light sensor. And the designed light sensor can work in the power down mode, which means when the light sensor is out of work, the power dissipation can decrease to almost zero. In the designed light sensor there is also an oscillator. The output frequency is 327.6KHz. The whole current of the analog circuits is about 330μA.