Hello everyone!

I made design of 5 x 5 TEM horns planar array in HFSS, and found that there are the great back lobes (comparable -5, -6 dB to the main lobe) in all most frequency range. Element period in array < lambda (min). Has anyone seen this problem before? and how to fix it? I don't know where the back lobes come from.
Thanks for advance!