| With the rapid development of noise reduction technology, noise-reducing headphone has become more and more popular among civil people. Though the theory of active noise reduction they have grasped is the same, the technology they have mastered is of unevenness, which leads to the diversity of the effect of noise reducing. Designing a thorough noise reduction testing system which could provide each kind of reduction data has become a hot topic in many companies and laboratories. Among many factors which affect the result of testing a headphone, the ambient noise is most important. Anechoic chamber can effectively solve the problem, but it can’t be widely used in schools and companies because of its high cost, large space occupation and poor portability, etc. Therefore, designing a testing system in complex noise environment is of great significance.The research and design on such a system home and abroad all concentrate on "reducing" the environmental noise interference, of which the specific methods all about to mantle the testing jig with a noise isolation tank or, arithmetically, calculating the frequency spectrum of the environmental noise with some kind of certain weight value on frequency spectrum and make it as a compensation for the final noise reduction data. And the testing method they have used is:firstly, measure the frequency response of the microphone without headphone and use it as a standard for calculating noise reduction; then measure the passive noise reduction and overall noise reduction with headphone separately. But such system proved to be out of people’s expectations, and the noise isolation tank would bring much trouble to the testing process. Relatively stability of environmental noise is the prerequisite of the method noise compensation. If changed, the weighting value must be calculated again.In view of this,we have made a breakthrough on testing method and design a new testing system, which could totally eliminate the influence of environmental noise. The system has 2 standard microphones and 2 artificial ears. Firstly, normalization without headphone is based on 2 channels audio sampling system, from which we can get the difference of the 2 channels’frequency response. Because the difference is determined only by the characters of the microphones and artificial ears, it could use as a compensation for the microphones. Compensated frequency response of microphones is considered to be that of artificial ears. Then the response of the microphones with compensation will be a standard for the calculation for noise reduction. The substance of this method is:treat environmental noise as useful signal and as a part of testing signal, or that the standard among every step of testing varies as the environmental noise varies. Imagine that if the deviation of the frequency response difference between microphones and artificial ears could be 0, then, theoretically, our system would own an ability of absolutely anti-noise performance. Among the process of designing our system, we have solved many problems, listed as follows:testing source signal design, signal sampling and storage, WAV player designing, dispose of sampling delay and data optimization, FFT and filter processing for signal, designing of 1/n octave filter.The system we offered contains two parts:software and hardware. The evaluation of our system includes anti-noise performance, accuracy and stability. Experimental methods are:contrast experiment in different environmental noise; comparing two sets of data from successive testing; testing without headphone and checking the distance between the reduction data and zero; testing the same headphone on our system and B&K PULSE, then comparing the data. According to the comparison of the testing data, the system is of great performance on anti-noise capacity and it can also offer all kinds of noise reduction data objectively which is used as gist for predication of the tested headphone that whether it can meet the requirement. |
PDF Full Text Request