Well this has quite a long title, lets explain the situation I'm currently working on the design of a sensor to measure the dielectric permittivity of a liquid medium, the idea is to use a ring oscillator and a delay line (the mentioned transmission line)
problems arise however due to the transmission line effects, the bandwidth of the drivers is around 750mhz (the actual oscillation frequency should be around 70-150M) and the separation between the conductors in the line must be reasonably high so as the fields to propagate outside the pcb and actually change the propagation time, in turn changing the frequency of the oscillator
the problem in question is that the impedance of the given wave guide actually changes when the medium changes, so the termination will be only suitable for a given range of dielectric permittivity, then it will start bouncing significantly, that wouldn't be much of an issue except that if the signal bounces from the tx side into rx with enough strength it can flip the inverter and the frequency would not be linear with propagation time anymore
so the question would be, is there some way of terminating the line that somehow compensates for this?
and also what about the transition zone from narrow to wide wave guide? there is not enough space to make a long tapered adapter, what method do you suggest to minimize reflection there?