Embedded Signal Processing Platform Application To Rotation Sensing System

Sensing technology is one of the three major issues in the Information Industry. Core sensing system consists of sensor head and signal data retrieving, the formoer responding to the physical quantity before transferring them into electronic quantity, while the latter in charge of changing the former into information we learn directly. Meantimewhile embedded-platform-based advanced microprocessor and real-time operating system is being converged into various applications of sensing system.The measurement theory of traditional non-contact tachometry focuses on mechanically and electronically counting the light pulses from rotating object. Thanks to the power of embedded platform, this paper comes up with the idea of adopting spectral analysis in tachometry realization. According to Fourier's theory, the application of spectral analysis to the periodical light pulse easily presents the RPM information we need. With ARM7-cored embedded processor LPC2210 and real-time operating system uC/OS-II together implementing the embedded platform, in which multi-task scheduling and Fast Fourier Transform being taken charge of, tachometry function is successfully performed.The use of emitting-detection-integrated optical sensor HBCS-1100 makes the whole system compact and flexible. uC/OS-II is ported into ARM7-cored LPC2210, FFT algorithm studied whilst codec provided, and other cooperative tasks coded to make it a whole real-time operating system prototype. Comparing the FFT results with that from Matlab conforms the accountability of our system, a set of experiments are carried out for the real object rotation sensing, after which lots of corresponding data acquired to plot the linear relationship between the processing result on the embedded platform and that of a hand tachometry.