Page 471 - Engineering Digital Design
P. 471

10.7 THE D FLIP-FLOPS: GENERAL                                      441


                                                                   State
                                                                  variable Input logic
                                                        State     change  value
                                                       variable, Q


                                                                          0
                             0 Reset
                                                                          1 Set
                             1 Set
                                                                 1 -> 0   0
                                                      u iT
                        Operation               ^ 1 J t          1 —»• 1   1 Set Hold
                          Table
                                                                   Excitation
                           (a)                    D                  Table
                                                 State                (c)
                                                Diagram
                                                  (b)
                  FIGURE 10.23
                  Generalized D flip-flop definition expressed in terms of the operation table (a), the state diagram (b),
                  and the excitation table (c).


                  diagram. For example, the Reset hold branching path 0 —> 0 is assigned D = 0 (for D),
                  and the Set branching path 0 —> 1 is assigned the D = 1 for branching condition D. The
                  excitation table for the D-FF is extremely important to the design of other state machines,
                  including other flip-flops, as will be demonstrated in later sections.
                    Now that the foundation for flip-flop design has been established, it is necessary to
                  consider specific types of D flip-flops. There are three types to be considered: the D-latch,
                  the edge triggered (ET) D flip-flop, and the master-slave (MS) D flip-flop, all of which
                  adhere to the generalized definition of a D flip-flop expressed in Fig. 10.23. Each of these
                  D-type flip-flops is represented by a unique state diagram containing the enabling input
                  clock (CK) in such a way as to identify the triggering mechanism and character of the D
                  flip-flop type. In each case the memory element used for the design of the D flip-flop is the
                  basic cell (set-dominant or reset-dominant) that is characterized by the combined excitation
                  table given in Fig. 10.15c. The design procedure follows that given in Section 10.6 where
                  use is made of the important mapping algorithm.

                  10.7.1 TheD-Latch

                  A flip-flop whose sequential behavior conforms to the state diagram presented in Fig. 10.24a
                  is called an RET transparent (high) D latch or simply D latch. Under normal flip-flop action
                  the RET D latch behaves according to the operation table in Fig. 10.23a, but only when
                  enabled by CK. The transparency effect occurs when CK is active (CK = 1). During this
                 time Q goes active when D is active, and Q goes inactive when D is inactive — that is,
                  Q tracks D when CK = 1. Under this transparency condition, data (or noise) on the D
                 input is passed directly to the output and normal flip-flop action (regulated by CK) does not
                  occur. If the D latch is itself to be used as a memory element in the design of a synchronous
                 FSM, the transparent effect must be avoided. This can be accomplished by using a pulse
                 narrowing circuit of the type discussed later. The idea here is that minimizing the active
   466   467   468   469   470   471   472   473   474   475   476