Matching Network and reflection coefficient

Thread Starter


Joined Apr 22, 2019
Hello there, I have a question about the design of matching networks. Consider for instance the output terminals of a radiofrequency amplifier: in order to avoid reflection on the load (which is a generic impedance), we may put between the amplifier and the load a matching network. For instance, consider a matching network made of lumped elements (precisely a T section), like in the following schemes (in the first picture, consider only the output matching network):

where the output matching network is:

About this one, my question is the following. The design procedure consists of choosing the values of L1, L2, C in order to get, at a certain frequency, that the total impedance (which depends on L1, L2, C and RL) at right is equal to Ropt conjugate (where with Ropt we refer to the output impedance of the amplifier). In this way there is not reflection at the output port of the amplifier. But, who guarantees me that there is not any reflection inside the matching network? Since inside it there is a variation of impedance (it converts Ropt to RL), I think there should be reflection.

In other terms, who guartantees that if I cut the matghing network (for example I consider the node between L1 and L2), the impedance seen from right is the complex conjugate of that seen from left? Why is this equality condition not requested in the design procedure?