The feed to a rod or wire dipole antenna can be well approximated by a discrete port if the fields at the feed are concentrated in a region that is small compared to a wavelength. If the feed region becomes significant compared to a wavelength, such as with microstrip feeding a planar dipole, then the impedance seen by a waveguide port feeding the microstrip will be different to the impedance seen by a discrete port. The fields around a discrete port contain a lot of short wavelength modes that die out very quickly at small distances from the port but contribute to the reactive impedance. The waveguide port does not generate these modes, but only the ones that propagate in the transmission line it is launching into. In CST it is necessary that this transmission line has a short distance from the port over which the transmission line is uniform, which prevents a waveguide port being used directly at the terminals of a dipole antenna and other structures not starting with uniform transmission lines. Presumably this is needed because of the way the code calculates or launches the waveguide modes.

Malcolm White thank you for your valuable answer.

Hello,

In CST, different ports are used for excitation to steer beams and produce maximum beams. Here's why:

• Phase anglesDriving each element with a different phase angle changes the radiation property. The phase shift is controlled so that the radiation adds up in a certain direction to produce a maximum beam.
• Beam movementControlling the phase angles in different ports allows the beam to be moved. This is the concept behind phased array.

Here are some types of ports available in CST:

• Discrete portsThese ports can be used to simulate lumped element sources. They can be voltage, current, or S-Parameter ports.
• Waveguide portsThese ports simulate an infinitely long waveguide, such as a coaxial cable, a microstrip line, or a rectangular waveguide. They provide a better match to the mode pattern and higher accuracy in S-Parameters.

Thanks,