Specification for radio disturbance and immunity measuring apparatus and methods - Part 4-3: Uncertainties, statistics and limit modelling - Statistical considerations in the determination of EMC compliance of mass-produced products
This part of CISPR 16 deals with statistical considerations in the determination of EMC
compliance of mass-produced products.
The reasons for such statistical considerations are:
a) that the abatement of interference aims that the majority of the appliances to be approved
shall not cause interference;
b) that the CISPR limits should be suitable for the purpose of type approval of massproduced
appliances as well as approval of single-produced appliances;
c) that to ensure compliance of mass-produced appliances with the CISPR limits, statistical
techniques have to be applied;
d) that it is important for international trade that the limits shall be interpreted in the same
way in every country;
e) that the National Committees of the IEC which collaborate in the work of the CISPR should
seek to secure the agreement of the competent authorities in their countries.
Therefore, this part of CISPR 16 specifies requirements and provides guidance based on
statistical techniques. EMC compliance of mass-produced appliances should be based on the
application of statistical techniques that must reassure the consumer, with an 80 % degree of
confidence, that 80 % of the appliances of a type being investigated comply with the emission
or immunity requirements. Clause 4 gives some general requirements for this so-called
80 %/80 % rule. Clause 5 gives more specific requirements for the application of the
80 %/80 % rule to emission tests. Clause 6 gives guidance on the application of the CISPR
80 %/80 % rule to immunity tests. The 80 %/80 % rule protects the consumer from noncompliant
appliances, but it says hardly anything about the probability that a batch of
appliances from which the sample has been taken will be accepted. This acceptance
probability is very important to the manufacturer. In Annex A, more information is given on
acceptance probability (manufacturer’s risk).