The significance of experimental results obtained with couplers designed for high directivity is addressed. Two aspects are considered. First, the effect of calibration errors on the coupler directivity is investigated and the related accuracy limits are formulated on a statistical basis. Second, the directivity degradation due to fabrication tolerances is predicted by simulation and verified experimentally. This allows to relate measurement results to theoretical expectations and to draw some more general conclusions on the relative importance of these two effects.
