Page 397 - Maxwell House
P. 397

DISCONTINUITY IN FEED LINES                                             377


            Pay attention to the asymmetrical definition of input and output column vectors that slightly
            simplifies the matrix element definition (see footnote below). Evidently, the expression (7.16)
            means that

                                      = ( ⁄
                          =   +    � ���������� �  1  12   )  + ( −   ⁄  )  2    (7.17)
                   1
                               12 2
                                                                            22
                        11 2
                                                        22
                                                                      21 12
                                                            1
                                                                 11
                   =   +     = (1  )  − ( ⁄  ) 
                                                    ⁄
                        21 2
                                                                       2
                               22 2
                                                                   22
                                                               21
                                                      22
                                                          1
                                               2
                   1
            Looking back at the expression (7.6), we can come to conclusion that the components of T- and
            S-matrix are connected as
                                                             ⁄
                             11  =  ⁄  ;    12  =  −   ⁄  ;   21  = 1  ;  22  = − ⁄         (7.18)
                                            21 12
                                                               22
                                                   22
                          12
                              22
                                       11
                                                                          21
                                                                             22
            Therefore,
                                                                           ⁄
                     11  =  12  −   ⁄  ;   12  =  ⁄  ;  21  = − ⁄  ;  22  = 1          (7.19)
                                                              22
                               11 22
                                     21
                                                                             21
                                               11
                                                                 21
                                                   21
            Notice  that the elements of  T-matrix  might reach  enormous  values (as  | | → 0  for high
                                                                          21
            reflective or extremely lossy circuit) in contrast to S-matrix. Clearly, such nuisance might create
            some difficulties for numerical analysis based on T-matrix. Taking into account (7.15), the
            reciprocal and lossless 2-port network can be described as
                                                        − 11  − 2 11  
                        Symmetrical, i.e.  22  =   and   =  �  11 �
                                            11
                                                      �1−| 11 | 2  − 11  1
                                                          − 11    2 11  
                                Asymmetrical, i.e.  22  = −  and  =  �  11 �               (7.20)
                                               11
                                                        �1−| 11 | 2   11  1
            Now assume that only two networks A and B are connected as Figure 7.3.3b demonstrates.
            Then
                                    2            2
                           1
                         � � = �  11  12 � �  �  and  �  1 � = �  11  12 � �  �                     (7.21)
                                                                 
                                                             
                           1   21   22   2   1   21   22   2
                                      
                                  
            Meanwhile, as soon as two networks are connected by the line of the same impedance,  2  =
                                        
                                                   
              and  2  =  1 4 , i.e. ( ,  ) = ( ,  ) , and thus
                                                 1
                                     2
                                             1
             1
                                 2
                                                      2
                                    1
                                                         � � = �  11  12 � �  11  12 � �  �                                    (7.22)
                                                          
                                                
                                           
                                                     
                                    1   21   22   21   22   2
            Repeating this procedure, we obtain for T-matrix of the cascade network in Figure 7.3.3a
                                                                      =  ⋅  ⋅  …                                                  (7.23)
                                                     3
                                                         
                                             1
                                                 2
            Therefore, the scattering transfer matrix  of the composite network is just the matrix product
            of the scattering transfer matrices of the networks in cascade. We shall see later (the expression
            (7.33)) that the straightforward S-matrix transformation for the cascade network is possible but
            not so trivial. Nevertheless, it can be done for any practically possible combination of networks
            while the T-matrix technique is not so powerful. Notably, the connected in cascade networks
            must have the equal number of input and output ports. Besides, the physical interpretation of
            T-matrix elements is a little bit vague. According to (7.19) only coefficient   has the clear
                                                                           22

            4  Now it became clear why we use the asymmetrical definition of input and output vectors
   392   393   394   395   396   397   398   399   400   401   402