Generation and effect of surface waves at a waveguide-ferrite boundary

A recent analysis by Gerosa of this problem, in which a wall-conductivity surface wave is incorporated, is shown to be in error in that the amplitude of the surface wave is undetermined by the formulation. The raison d´Ãªtre of this surface wave is examined here, and it is shown to be needed to prevent a nonphysical divergence of the junction magnetic field which would occur in just the same parameter range as that in which the surface wave exists. Despite the necessity for the existence of this surface wave, it is shown to be of negligible proportions as the wall conductivity increases, and in the limit it does not contribute to the junction discontinuity impedance.