Making Circular Polarized Sources for Analysis

Before hiring time on kit that runs HFSS, AWR or CST, etc, it would be handy to know..

Can the port that is used to feed a circular polarized structure, say like a circular section
feed-horn, be a simple linear polarized port where we simply expect 3dB loss?

Can one stack drive ports on the same spot, with the polarization directions at
90 degrees to each other, and one of them time delayed by a quarter wavelength?

Might it be necessary to place one wave port at the feed point of the horn, and
place a second port a quarter wavelength back, on the end of a short conducting
tube, and oriented to be at 90 degrees to the first port as a handy dodge to make
circular polarized waves