Even/odd mode velocity vs. bandwidth
From a time domain view, I am speculating that if two modes travel at different speeds, then the added signal gets distorted. But I feel this thought is still clear enough.
Thank you.
An even/odd mode velocity disparity, like you get in microstrip coupled structures, means literally that some of the energy will arrive at the output by one path, and some of the energy will arrive at the output by another path (in a mathematical superposition sort of way). There, the two will vectorially interact (add or subtract) from each other causing gain ripple.
So, in order to get broadband performance, the problem is like trying to design TWO parallel circuits that have the same phase characteristics. In a TEM structure, you only need to design one network to do the same--easier to do!
biff44, thanks for the participation.
I don't think the problems (bandwidth limitation or gain ripple you mentioned) are due to the co-existance of even and odd modes. It is due to the difference in their phase velocities instead. Please note that even for a TEM coupled transmission line, such as coupled stripline, there are still even and odd modes. It is just that they have identical velocities (at least theoretically).
This is now off the topic a little, since originally I was trying to get a good answer for why the difference in velocities causes bandwidth reduction.
