Research Of Noise Detecting System For Scroll Compressor Based On Vibration Signal

Scroll compressor is widely used in air conditioners, refrigerators, refrigeration and other electrical equipment. As the equipment of the main noise source, reducing the noise in the production of the manufacturing sector has been paid more and more attention by the manufacturers. We can understand the noise level, source and its distribution through testing the compressor, so detection is an essential part of achieving noise reduction.As for the compressor noise measurement, the most current common measure is to move the compressor to anechoic chamber for sampling inspection, and this is obviously not conducive to the complicated practical production because of harsh experimental conditions and testing procedures. Aimed at these shortcomings existed in traditional test, a new compressor noise detection method is put forward based the relationship vibration and noise. According to the correlation between sound radiation and vibration and noise radiation, sound radiation efficiency curve is obtained via experiment, and then we can simplify the measuring points through observing and analyzing the distribution of sound intensity. This method can accurately reflect the whole noise status of compressor by just collecting some field vibration signals on some certain measuring points. Compared to traditional method, this method is easy and simple, and has an important guiding significance for actual production.Two main parts of contents are contained in this research including the theory of reflecting noise using vibration and the development of noise detection system. For the former, the basic acoustics and the basic theory and methods of sound intensity measurement is introduced firstly, and the structure, as well as the working principle and characteristics of scroll compressor is also analyzed. Secondly, the theory of reflecting noise using is investigated for further study and designs vibration and noise-related experiment program. Final test shows that it has a high correlation between vibration and noise and we can predict the noise via vibration signals. At the same time, we also get its sound radiation efficiency curve which laid the foundation for the development of noise detection system.According to the needs of noise detection, a noise detection system based embedded Linux is put forward in the research. Data collection is done through the audio chip UDA1341 controlled by IIS bus and vibration signals coming from 4-channel accelerometer sensor are amplified through acquisition module. And then, these signals are sent to S3C2410 for further processing after A/D conversion. Here, we uses DMA mode to save and the results are displayed on the LCD. The system not only contains data collection, time domain analysis, frequency domain analysis and others basic functions, but also 1/3 octave and noise decibel is obtained through processing using a constant percentage of the first digital filter for band-pass filter. Applied to practical application, it can be included that the developed noise detection system can collect data accurately, and has higher prediction accuracy, as well as good man-machine interface.