Page 584 - Engineering Digital Design
P. 584

554                CHAPTER 11 / SYNCHRONOUS FSM DESIGN CONSIDERATIONS


                            D flip-flops to be used for the synchronizer. Consider that at 200 MHz, it is
                            necessary to use a synchronizing scheme of the type shown in Fig. 11.23a for
                            which a divide-by-4 counter is used in the diagram.
                            (a) Calculate the MTBF (in seconds, days and years) if r — 0.5 ns and T 0 =
                                      6
                               1 x 10~  seconds. [Hint: Use Eq. (11.3) and take into account the cumulative
                               effects of both stages.]
                            (b) Repeat part (a) if a divide-by-two counter is used in the diagram. Calculate
                               the MTBF in seconds for comparison with the result in part a.

                     11.7   It is desired to find a safe operating clock frequency for a given FSM. The following
                            data is collected relative to the operation of the FSM:

                              Flip-flop parameters f pi h = 6 ns; f pM = 9 ns
                              Maximum delay through the NS logic T NS = 1 ns
                              Maximum flip-flop setup time t su = 3 ns

                            Calculate a safe operating clock frequency fcK based on a 15% safety factor.
                    11.8    Derive the expressions for Vx(t) in Eqs. (11.8) and (11.9) relative to Fig. 11.31c.
                            Assume that R\ ^> R2 and that the switch is opened (or closed) at time t = 0 only
                            after steady-state conditions are reached. State any simplifying assumptions that
                            are made relative to the Schmitt trigger and inverter. (Note: This exercise involves
                            solving a first-order RC circuit.)
                    11.9    Presented in Fig. PI 1.4 is the state diagram for a one-bit serial adder. The operand
                            bits, a and b, are introduced serially and are synchronized antiphase to the clock
                            triggering edge. The outputs are S (sum) and C 0 (carry-out). Assume that the FSM
                            is initialized (reset) after each addition operation.
                            (a) Complete the state diagram by giving it a state code assignment that is free of
                               ORGs.












                                                                         C IT












                    FIGURE P11.4
   579   580   581   582   583   584   585   586   587   588   589