This article investigates the risk to declare compliant a faulty equipment under test during a radiated susceptibility test performed in a reverberation chamber (RC) working imperfectly, i.e., in the lowest frequency range when the chamber is undermoded. The results obtained from an extensive measurement campaign are compared, using the same worst-case approach, to a probabilistic model proposed recently for well-stirred RCs. Below the frequency when the RC can be assumed to be well-stirred, the variance (or the standard deviation) of the electromagnetic field (or of the induced power on the equipment) may strongly differ from the ideal case, depending on the effect of the stirring process and of the modal density. If the variance is low, the test is, as expected, quite meaningless and it makes sense to avoid this corresponding frequency range. However, more surprisingly, if the variance is higher than the one observed in ideal conditions, the risk is paradoxically reduced for weakly susceptible equipment (but increased for highly susceptible ones). Indeed, this higher variance makes the appearance of a strong amplitude of the disturbance for a given stirring configuration more probable. This article is considered as an incremental step in the understanding of the RC behavior from the point of view of radiated susceptibility testing, particularly when the RC is undermoded, a frequency range generally used to perform this kind of tests.

中文翻译：

---

在低模混响室中进行辐射敏感性测试期间，有可能声明EMC符合故障EUT的风险

本文研究了在不完美工作的混响室（RC）中（即，当室未充分改造时的最低频率范围内）进行的辐射敏感性测试过程中，宣布合格的被测故障设备的风险。使用相同的最坏情况方法，将从广泛的测量活动中获得的结果与最近针对搅拌良好的RC提出的概率模型进行了比较。在可以假定RC充分搅拌的频率以下，根据以下因素的影响，电磁场（或设备上的感应功率）的方差（或标准偏差）可能与理想情况有很大差异。搅拌过程和模态密度 如果方差低，则测试如预期的那样毫无意义，因此避免该相应的频率范围是有意义的。但是，更令人惊讶的是，如果方差大于理想条件下观察到的方差，则对于弱势敏感设备的风险会自相矛盾地降低（但对高度敏感设备的风险会增加）。实际上，对于给定的搅拌配置，这种较高的变化使得出现较大的干扰幅度的可能性更大。从辐射敏感性测试的角度来看，本文被认为是理解RC行为的一个增量步骤，尤其是当RC调制不足时，通常用于执行此类测试的频率范围。对于给定的搅拌配置，这种较高的变化使得出现较大的干扰幅度的可能性更大。从辐射敏感性测试的角度来看，本文被认为是理解RC行为的一个增量步骤，尤其是当RC低调时，通常用于执行此类测试的频率范围。对于给定的搅拌配置，这种较高的变化使得出现较大的干扰幅度的可能性更大。从辐射敏感性测试的角度来看，本文被认为是理解RC行为的一个增量步骤，尤其是当RC低调时，通常用于执行此类测试的频率范围。