Study On Radiated Susceptibility Tests Using Mode-stirred Reverberation Chamber

Compared with the traditional test environment for EMC,the field in reverberation chamber(RC)is statistically uniform,isotropic and randomly polarized.Due to the high quality factor of the cavity,the smaller input power can form a large field strength,which also can simulate the actual electromagnetic environment for the equipment under test,and obviously RC is very suitable for radiated susceptibility test.For the radiated susceptibility test method with mode-stirred RC,it has higher test efficiency,and it has a broad application prospect for the aviation and military equipments,while the field variability is large,which may lead to greater uncertainty of tests.For IEC 61000-4-21:2011,there is no detailed discussion for the measurement uncetainty,and the uncertainty for the number of independent samples determined by the cyclic autocorrelation coefficient method is large.Moreover,for the radiated susceptibility test method,there are some problems such as the difficulty to determine and evaluate the speed of stirrer,as well as how to determine the maximum field.For this,the paper mainly concentrates on the radiated susceptibility tests using mode-stirred RC,the contents are as follows:1.The electromagnetic field of the cavity has been analyzed using the Green function method which can provide an analytical solution.Then,from the perspective of eigenmode superposition,the formal solution of the field strength at any point in the cavity has been analyzed.And for the RC,the plane wave superposition has been used to provide a theoretical basis for subsequent research.2.According to the problems in the existing standards,the new methods for evaluating the performance parameters of RC,e.g.,the number of independent samples,the uniformity of the mode stirred field,and the stirring speed,have been proposed.Aiming at the large uncertainty of the number of independent samples determined by the cyclic autocorrelation coefficient method,the method related to Pearson correlation coefficient is used to estimate the number of independent samples of mechanical stirrer.Based on the scattering parameter model of the RC,the Anderson-Darling(AD)goodness of fit(GOF)and the ergodicity test are used to evaluate the basic principles of the field distribution characteristics of the RC,and the scattering parameter samples are determined to assess the overmoded frequency.Aiming at the difficulty of evaluating the stirring speed in the mode stirred,a method using AD GOF to determine the overmoded frequency is proposed.And also,based on the Q value distribution,a method for evaluating the stirring speed using the coefficient of variability is proposed.For the two methods,they are all used to evaluate the stirring speed.The AD GOF test is used to analyze the influence of scanning time and stirring stroke on the field distribution in mode stirred.The field distribution is basically not affected by the scanning time;within the same stirring stroke,the field uniformity is not affected;but In the different stirring strokes,the stirrer alternately presents "concave" and "convex" structures during the stirring process,which significantly affects the field uniformity.3.Aiming at the problem that the maximum field determined by the extreme value distribution is affected by the number of independent samples,a statistical method using generalized extreme value(GEV)distribution to analyze the maximum field in RC is proposed.Increase independent sample groups through joint stirring or spatial sampling,and extract the maximum value from each independent samples to obtain sufficient maximums.Considering the sample size,for large samples with N ? 50,the parameters in GEV distribution are determined by maximum likelihood estimation.Conversely,L-moment estimation is used for small samples with N < 50.When the sample N ? 30,the normal distribution is used to determine the 95%confidence interval of the estimated parameters,and the t distribution is used for the small sample.Experiments have verified the statistical characteristics of the maximums.Since the oscillating wall stirrer will affect the mode shift,the maximums extracted by the stirring will have an impact on the characteristic frequencies such as undermoded,weakly-and highly-overmoded.A comparison of the maximum field in the RC determined by the extreme value distribution and the GEV distribution shows that the 95% confidence interval of the maximums is basically not affected by the sample size due to the short tail characteristics of the GEV.Based on this,a method to determine the input power required to form a certain field strength in RC based on the GEV theory is given,which provides effective technical support for scientifically constructing the radiated immunity/susceptibility tests.4.To assess the influence of the loading and its position in RC on mode-tuned/mode-stirred field distribution and maximum field,the bench type and floor type equipments under test are measured by loading in and out of the working volume respectively.Considering the quality factor Q and the mode-tuned/modestirred field distribution,loading at different positions will not cause significant effects.Loading outside the working volume can be used to improve the field caused by the "concave" structure.Different positions inside and outside the working volume loading will affect the maximum field formed in radiated susceptibility test.When the equipment under test is on the ground,the field effect near a perfectly conducting boundary of the RC should be considered.