Wave-Port Dimensions for Planar Folded-Dipole

I'm simulating a multiple fold dipole antenna for 200 GHz (lamda_g = 597.2 um). The antenna is folded 4 times (has 5 poles) for very high impedance. The antenna width and arm spacing is 0.02*lamda_g. The feed gap is 24.5 um. So I made the wave-port 24.5 by (s + w) um. Is this acceptable? Or is a lumped port more appropriate? The antenna is on Si that is about lamda_g thick and there is no ground plane. My radiation boundary is lamda_g/3 away from the antenna on all sides (dielectric side and air side). Should I increase the boundary distance to lamda_0/3 on the air side, and/or reduce the distance on the dielectric side to 1/3 of the wavelength in the dielectric (instead of at the dielectric air boundary)?

My file is attached. I just changed the file extension to *.exe so I could attach it. To open, just change it back to *.hfss. Thanks for any recommendations.

Here is an image of the antenna and wave-port. I just made it much wider. Now it extends across the other arms and not just the feed arm of the dipole. Also, when doing a port only solution, how do I view the E-field of the modes so I can check to make sure the correct modes are used for the wave port?