1 6 CHAPTER 6 DESIGN OF SEQUENTIAL LOGIC CIRCUITS IN QCA 6. INTRODUCTION The logic circuits whose outputs at any instant of time depend not only on the present inputs but also on the past outputs are known as sequential circuits. In sequential circuit, outputs are fed back to the input side. Thus an output signal is a function of the present input signals and a sequence of the past input signals. The block diagram of sequential circuit is shown in Figure 6.. It consists of a combinational circuit to which memory elements are connected to form a feedback path. The storage elements are devices capable of storing binary information. The binary information stored in these elements at any given time defines the state of the sequential circuit at the time. Figure 6. Block Diagram of sequential circuit The sequential circuit can be classified into two types depending on the timing of their signals. They are
2 7 ) Synchronous or clocked sequential circuits 2) Asynchronous or unclocked sequential circuits. In the synchronous sequential circuit, synchronization is achieved by a timing device called a master clock generator which generates a periodic train of clock pulses. In practice, it may be achieved by applying clock pulses to various AND gates through which external inputs enter the sequential circuit. This ensures the gates to transmit input signals only which coincide with the arrival of the clock pulses. In an asynchronous sequential circuit, events can occur after one event is completed and there is no need to wait for a clock pulse. Therefore, in general, asynchronous circuits are considerably faster than the synchronous sequential circuits. However, in an asynchronous circuit, events are allowed to occur without any synchronization. In this thesis, design of sequential circuits such as Flip Flops (FF), Shift Registers (SR) and Counters are proposed. The different sequential circuits are designed by using majority gate. Then the designed circuits are implemented with QCA cells and simulated using QCADesigner. These structures are popular designs in the transistor technology. The circuit designs in QCA follow the conventional design approaches, but due to the technology differences, they are modified and optimized for the best performance in QCA. 6.2 DESIGN OF FLIP FLOPS (FF) USING MG The memory elements in a sequential circuit are called flip-flops. A circuit has two outputs, one for the normal value and one for the complement value of the stored bit. Binary information can enter a flip-flop in a variety of
3 8 ways and gives rise to different types of flip-flops. The most common types of flip flops are. SR (Set-Reset) Flip Flop 2. D (Data or Delay) Flip Flop 3. T (Toggle) Flip Flop 4. JK Flip Flop The flip flops are synchronous sequential circuits. Hence the circuits change their states only when clock pulses are present. The basic operation of these flip flops has been explained by Morris Mano (995). This section describes the architectures for the four flip-flops. By replacing AND/OR logic with majority logic it is possible to significantly minimize the total number of cells in the design. Since there is no static memory equivalent in QCA, hence to create loops, which keep memory in motion. The proposed architectures of the flip-flops have been shown below. As a means for generating the QCA architecture an obvious first step is to translate existing CMOS designs directly into QCA majority gate logic. However, while such a translation is possible, the nature of QCA devices will require an architecture that is radically different from conventional CMOS. Hence first the sequential circuits are designed by using conventional gates. Then it can be translated into QCA majority gate Design of SR Flip Flop The clocked SR flip-flop shown in Figure 6.2 consists of a basic NOR flip-flop and two AND gates. The outputs of the two AND gates remain at as long as the clock pulse (or CP) is, regardless of the S and R input values. When the clock pulse goes to, information from the S and R inputs passes through to the basic flip-flop. With both S= and R=, the occurrence of a clock pulse causes both outputs to momentarily go to. When the pulse
4 9 is removed, the state of the flip-flop is indeterminate, i.e., either state may result, depending on whether the set or reset input of the flip-flop remains a longer than the transition to at the end of the pulse. (a) Logic diagram (b) Graphic symbol Figure 6.2 SR flip flop Table 6. Characteristic table of SR flip flop Q S R Q(t+) Intermediate Intermediate Figure 6.3 Characteristic equation Q (t+) = S+R'Q (6.)
5 2 The characteristics table of the flip flop is shown in Table 6.. This table shows the operation of the flip flop in a tabular form. The Q is an abbreviation of Q (t) and stands for the binary state of the flip flop before the application of a clock pulse, referred to as the present state. The S and R columns give the possible values of the inputs, and Q (t+) is the state of the flip flop after the application of a single pulse, referred to as the next state. The table must be interpreted as follows: Given the present state Q and the inputs S and R, the application of a single pulse in the CP input causes the flip flop to go to the next state Q (t+). The characteristic equation of the flip flop is derived by using the map of Figure 6.3. This Equation (6.) specifies the value of the next state as a function of the present state and the inputs. The characteristic equation is an algebraic expression for the binary information of the characteristic Table 6.. The two intermediate states are marked with don t-care X s in the map, since they may result in either or. However, the relation SR = must be included as part of the characteristic equation to specify the both S and R cannot equal to simultaneously. Next the SR flip flop is designed using majority gates with the help of the characteristics equation. It is constructed by using 5 majority gates and inverter as shown in Figure 6.4. The circuit have two inputs S and R along with one control input Clock. The last two majority gates M4 and M5 form a loop. The output of the flip flop is fed back to M4 and combined with the input R'of majority gate M4 to produce R'Q. Here the gate M4 act as an AND gate. The gate M5 is designed to perform OR operation which combine the output of Majority gate M and M4.Hence the output of the gate M5 produce the desired characteristic Equation (6.).
6 2 Figure 6.4 Design of SR flip flop using majority gate Computation with QCA is accomplished by designing QCA layouts, which exhibit the desired interaction of states. The designed SR flip flop using majority gate is next implemented by using QCA cells. The flip flop is implemented by 57 cells, with an area of 72, nm 2. The cell minimization techniques are used to minimize the number of cells from 78 to 57. The corresponding QCA implementation of SR flip flop is shown in Figure 6.5. Figure 6.5 Layout of SR flip flop Design of D Flip Flop One way to eliminate the undesirable condition of the indeterminate state in the RS flip flop is to ensure that inputs S and R are
7 22 never equal to at the same time. This is done in the D flip flop as shown in Figure 6.6 (a). The D flip flop is a modification of the clocked SR flip-flop. It has only two inputs: D and CP. The D input goes directly into the S input and the complement of the D input goes to the R input. The D input is sampled during the occurrence of a clock pulse. If it is, the flip-flop is switched to the set state (unless it was already set). If it is, the flip-flop switches to the clear state. (a) Logic diagram (b) Graphic symbol Figure 6.6 D flip flop Table 6.2 Characteristic table of D flip flop Q D Q (t+) Figure 6.7 Characteristic equation Q (t+) = D (6.2)
8 23 The binary information present at the data input of the D flip flop is transferred to the Q output when the CP input is enabled. The output follows the data input as long as the pulse remains in its state. When the pulse goes to, the binary information that was present at the data input at the time the pulse transition occurred is retained at the Q output until the pulse input is enabled again. The characteristics table of the D flip flop is shown in Table 6.2. It shows that the next state of the D flip flop is independent of the present state since Q (t+) is equal to the input D whether Q is equal to or. This means that an input pulse will transfer the value of input D into the output of the flip flop independent of the value of the output before the pulse was applied. The characteristics equation shows clearly that Q (t+) = D. Next the D flip flop is designed with the help of the characteristic equation. The flip flop is designed by two ways. First method is using majority gates and the second method is using a QCA binary wire. In the first method the flip flop is constructed by using 3 majority gates and inverter as shown in Figure 6.8. The circuit have one input D and along with one control input Clock. The output Q of the flip flop is fed back to the majority gate M3 and combined with the input D to produce the output Q. Here the gate M act as an AND gate and M2 act as an OR gate. The gate M3 is designed to perform OR operation which combine the output of Majority gate M and M2.Hence the output of the gate M3 produce the desired characteristic Equation (6.2).
9 24 Figure 6.8 Design of D flip flop using majority gate The QCA implementation of the D flip flop is shown in Figure 6.9. The QCA implementation requires 69 cells, with an area of 93,686 nm 2 and this also required less number of cells than previous implementations. This is achieved by using cell minimization techniques. Figure 6.9 Layout of D flip flop Latching is effectively accomplished through timing by using a four-phase clocking arrangement. Therefore, a D-type FF can be constructed by a QCA binary wire with four clocking zones as shown in Figure 6.. In this case, the input signal is delivered to the output after at least one complete clock cycle delay and control is accomplished by timing. The relative simplicity of a D-FF seems to suggest that sequential design in QCA could be achieved at ease within the Cartesian layout and have been proposed by
10 25 Huang et al (27). However, timing and signal delay must be carefully considered. Figure 6. Layout of D flip flop using binary wire The wire which is clocked from left to right with increasing clocking zones will carry information in the same direction. The clock signals act to pump information in the circuit as a result of the successive latching and unlatching in cells connected to different clock phases. Figure 6. shows a wire connected to different clock zones. Each group of cells connected to a particular clocking zone. The clocked QCA wire is used to carry the information which is constructed by regular QCA cells. Hence in the QCA wire input and outputs are equal and it act as a D flips flop. The D flip flop is implemented by only 6 cells Design of JK Flip Flop A JK flip-flop is a refinement of the SR flip-flop in that the indeterminate state of the SR type is defined in the JK type. Inputs J and K behave like inputs S and R to set and clear the flip-flop (note that in a JK flipflop, the letter J is for set and the letter K is for clear). When logic inputs are applied to both J and K simultaneously, the flip-flop switches to its complement state, i.e., if Q=, it switches to Q= and vice versa. The Output Q is ANDed with K and CP inputs so that the flip-flop is cleared during a clock pulse only if Q was previously. Similarly, output Q' is ANDed with J and CP inputs so that the flip-flop is set with a clock
11 26 pulse only if Q' was previously. The behavior of the JK flip flop is demonstrated in the characteristics table. A clocked JK flip-flop is shown in Figure 6.. It is very important to realize that because of feedback connection in the JK flip-flop, a CP signal which remains while both J and K are equal to will cause the output to complement again and repeat complementing until the pulse goes back to zero. To avoid this, the clock pulses must have time duration less than the propagation delay through the flip-flop. The restriction on the pulse width can be eliminated with a master-slave or edgetriggered construction. (a) Logic diagram (b) Graphic symbol Figure 6. JK flip flop Table 6.3 Characteristic table of JK flip flop Q J K Q(t+)
12 27 Figure 6.2 Characteristic Equation Q (t+) = JQ'+K'Q (6.3) Next the JK flip flop is designed by using the characteristic equation. The flip flop is constructed by using 4 majority gates and 2 inverters as shown in Figure 6.3. The circuit has two inputs J and K along with one control input clock. The output Q of the flip flop is fed back to the input majority gate M and the complement of the output is fed back to the majority gate M3. The output of M and M3 are combined with the clock input by M2. Here the gate M2 act as an OR gate to produce JQ +K Q. The gate M4 is used to form a loop and to produce the desired characteristic Equation (6.3). Figure 6.3 Design of JK flip flop using majority gate
13 28 The QCA implementation of the JK flip flop is shown in Figure 6.4. The QCA implementation requires 46 cells, with an area of 36, nm 2. In the proposed QCA implementation, the required number of cells is 5% reduced from the previous implementations. This is achieved by using cell minimization techniques. Anoop Vetteth et al (23) have proposed QCA architecture for the JK flip-flop. It consists of 7 majority gates and takes 2 / 4 clock cycles to complete the output, while our new QCA architecture requires only one clock cycle. Figure 6.4 Layout of JK flip flop Design of T Flip Flop The T flip-flop is a single input version of the JK flip-flop. As shown in Figure 6.5, the T flip-flop is obtained from the JK type if both inputs are tied together. The output of the T flip-flop "toggles" with each clock pulse. The designation T comes from the ability of the flip flop to toggle or complement, its state. Regardless of the present state, the flip flop complements its output when clock pulse occurs while input T =. The characteristics table and the characteristic equation show that when T =, Q (t+) = Q, that is, the next state is the same as the present state
14 29 and no change occurs. When T =, Q (t+) = Q', the state of the flip flop is complemented. (a) Logic diagram (b) Graphic symbol Figure 6.5 T flip flop Table 6.4 Characteristic table of T flip flop Q T Q (t+) Figure 6.6 Characteristic equation Q (t+) = TQ'+T'Q (6.4) Next the T flip flop is designed by using the characteristic equation. The flip flop is constructed by using 4 majority gates and 2 inverters as shown in Figure 6.7. The circuit has one input T and along with one control input
15 3 Clock. The output Q of the flip flop is fed back to the input majority gate M2 and the complement of the output is fed back to the majority gate M. The output of M and M2 are combined with the Clock input by M3. Here the gate M3 act as an OR gate to produce TQ'+T'Q. The gate M4 is used to form a loop and to produce the desired characteristic Equation (6.4). Figure 6.7 Design of T flip flop using majority gate The QCA implementation of the T flip flop is shown in Figure 6.8. The QCA implementation requires 59 cells, with an area of 6, nm 2. In this implementation, the required number of cells is 5% reduced from the previous implementations. This is achieved by using cell minimization techniques. Figure 6.8 Layout of T flip flop
16 3 The performance analysis proposed SR, D, JK and T flop flops are given in Table 6.5. The performance analyses of those circuits are compared according to the complexity, area, and number of clock cycles and the proposed designs are compared with existing majority gate method. Flip Flops Complexity (No. of QCA cells) Table 6.5 Comparison of Flip flops Previous structures Area Delay (No. of clock cycles) Complexity (No. of QCA cells) Proposed structures Area Delay (No. of clock cycles) SR 7 8cells 3nm x 36nm 2 57 cells 24nm x 3nm D 7 cells 36nm x 278nm 2 69 cells 337nm x 278nm JK 92 cells 32nm x 36nm 2 46 cells 2nm x 3nm T 97 cells 325nm x 38nm 2 59 cells 2nm x 3nm 6.3 DESIGN OF SHIFT REGISTERS IN QCA A register is a group of flip flops suitable for storing binary information. Each flip flop is a binary cell capable of storing one bit of information. An n-bit register has a group of n flip flops and is capable of storing any binary information containing n bits. The register is mainly used for storing and shifting binary data entered into it from an external source. It is a type of sequential circuit and an important building block used in digital systems like multipliers, dividers, memories, microprocessors, etc. A register capable of shifting binary information either to the right or to the left is called a shift register. The shift register permits the stored data to move from a particular location to some other location within the register. In a shift register, the FFs are connected in such a way that the bits of a binary number are entered into the shift register, shifted from one position to another and finally shifted out.
17 32 There are two methods of shifting the data. They are ) Serial shifting and 2) Parallel shifting. The serial shifting shifts one bit at a time for each clock pulse in a serial fashion, beginning with either MSB (Most Significant Bit ) or LSB (Least Significant Bit)For example, a 4-bit register requires four clock pulses to shift a bit from the input to the output. In parallel shifting operation, all the data get shifted simultaneously during a single clock pulse. Hence it is faster than the serial shifting method. These two methods can be used to shift data into a register and out of the register. Shift registers are classified into the following four types based on how binary information is entered or shifted out. They are. Serial in Serial out (SISO), 2.Serial in Parallel out (SIPO) 3. Parallel in Serial out (PISO) and 4.Parallel in Parallel out (PIPO) The block diagrams of the four basic register types are shown in Figure 6.9. The registers can be designed using flip flops (Morris Mano 995). An n-bit shift register consists of n-flip flops and the required gates to control the shift operation. Figure 6.9 Shift register types
18 33 Shift registers are used in digital systems for temporary storage of information, data manipulation and transferring. In addition, they are used in counting circuits, such as simple counters, variable modulo counters and increment counters. In this thesis, the Serial-in Serial-Out (SISO) and Serial-in Parallel- Out (SIPO) 3-bit shift registers are designed using D and JK flip flops which allow shifting in both directions: Left to Right or Right to Left. The results are verified according to the truth table. The performance analyses of those circuits are compared according to the complexity, area, and number of clock cycles Serial in Serial Out Shift Register This type of shift register accepts data serially, i.e. bit at a time on a single input line. It produces the stored information on its single output also in serial form. Delay data by one clock time for each stage. It will store a bit of data for each register. Data may be shifted left using shift left register or shifted right using shift right register Shift right register A shift right register can be built using JK flip flops or D flip flops as shown in Figure 6.2 and Figure 6.2 respectively. A JK flip flop based shift register requires connection of both J and K inputs. Input data are connected to the J and K inputs of the rightmost (lowest order) flip flop. To input a, i.e., J = and K = ; to input a, a at J input, i.e., J = and K =, should be applied. When a clock pulse is applied, the data will be shifted bit by bit to the right.
19 34 Figure 6.2 SISO Shift right register using JK flip flop Figure 6.2 SISO Shift right register using D flip flop In the shift register using D flip flops; D input of the leftmost flip flop is used as a serial input line. To input data, one should apply a at D input and to input data, a at the D input should be applied. Figure 6.22 Data shift in SISO Shift register For example, consider that all stages are reset and a steady logical is applied at the serial input line connected to stage A. A is applied at the serial input line i.e., at D input of the first flip flop (right most). When the first clock pulse is applied, flip flop A is SET, thus storing the. Next a is applied to the serial input, making D= for flip flop A and D= for flip flop B because D input of B is connected to the Q A output.
20 35 When second clock pulse occurs, the on the data input is shifted to flip flop A and in flip flop A is shifted to flip flop B. The next in the binary number is now applied at the serial line and a clock pulse is applied. This is entered into flip flop A and the stored in flip flop A is shifted to flip flop B and the stored in flip flop B is shifted to flip flop C. This completes the serial entry of the 3-bit binary number into the shift register. The data in each stage after each of the three shift pulses is shown in Table 6.6. The logical- input enters into stage A and then shifts right to stage C after 3clock pulses. This is graphically explained in Figure For a JK FF, the data bit to be shifted into the flip flop must be present at the J and K inputs when the clock transition from low to high occurs. Since the data bit is either a or a, there are two cases:. To shift a into the FF, J= and K=and 2. To shift a into the FF, J= and K=. Table 6.6 Truth table of SISO shift right register QCA implementation of Shift right register The shift right register is constructed by using 3 flip flops (JK flip flops or D flip flops).the flip flops are designed by using majority gates as explained in previous section. The shift register is implemented by QCA cells using flip flops implementation in section 6.2. It is a 3 bit shift register hence 3 flip flops are used in the design.
21 36 In the JK flip flop design as shown in Figure 6.23, the shift register is constructed by using 28 cells, with an area of 267,48 nm 2. The design of each flip flop involves 4 clocking zones i.e, one clock cycle. Hence inputoutput delay for the shift register design required 3 clock cycles. In the shift register using D flip flop as shown Figure6.24, design involves only 3 cells and delay of 3 clock cycles. The area required for this QCA design is 3,64 nm 2. Figure 6.23 Layout SISO Shift right register using JK flip flop Figure 6.24 Layout SISO Shift right register using D flip flop Shift left register A shift left register can be built using JK flip flops and D flip flops as shown in Figure 6.25 and Figure 6.26 respectively. The clock pulse is applied to all the flip flops simultaneously. When the shift or clock pulse occurs, each flip flop is set or reset according to the data at the respective flip flop input. Thus, the input data bit at serial input line is entered into stage A
22 37 by the first clock pulse. At the same time, the data of stage A is shifted into stage B and so on for the following stages. For each shift pulse, data stored in the register stage shifts to the left by one stage. New data is entered in to stage A, whereas the data present in stage C are shifted out (to the left) for use by some other shift register or other building block of digital system. Figure 6.25 Schematic of SISO Shift left register using D flip flop Figure 6.26 Schematic of SISO Shift left register using JK flip flop For example, consider that all stages are reset and a steady logical is applied at the serial input line connected to stage A. The data in each stage after each of the three shift pulses is shown in Table 6.7. The logical input enters into stage A and then shifts left to stage C after 3clock pulses. Table 6.7 Truth table of SISO shift left register
23 QCA implementation of Shift Left register The shift left register can be implemented by QCA cells using JK flip flops or D flip flops as shown in Figure 6.27 and Figure 6.28 respectively. This QCA design is similar to the QCA implementation of shift right register except the direction of data shifting. Hence there is no change in the number of cells required, area and clock cycles. Figure 6.27 Layout SISO Shift left register using JK flip flop Figure 6.28 Layout SISO Shift left register using D flip flop Serial in Parallel out Shift Register A 3-bit serial in parallel out shift registers are shown on Figure 29 and Figure 3. It is constructed by using JK flip flops and D flip flops. It consists of one serial input and outputs are taken from all the flip flops
24 39 parallel. In this register, data is shifted in serially but shifted in parallel, it is necessary to have all the data available at the outputs at the same time. Once the data is stored, each bit appears on its respective output line and all the bits are available simultaneously, rather than on a bit-by bit basis as with the serial output. Figure 6.29 Schematic of SISO Shift right register using D flip flop Figure 6.3 Schematic of SIPO Shift right register using JK flip flop QCA implementation of Serial in Parallel out Shift registers The shift register can be built by QCA cells using JK flip flops and D flip flops are shown in Figure 6.3 to Figure 6.34.The SIPO shift registers are implemented by QCA cells in both directions of data shifting. It consists of one serial input S through which the data is entered in to the register serially. The outputs A, B and C are taken from the flip flops in parallel. It is a 3 bit shift register hence 3 flip flops are used in the design.
25 4 In the JK flip flop design, the shift register is constructed by using 28 cells, with an area of 267,48 nm 2. This QCA design is similar to the QCA implementation of serial in serial out shift register. Hence there is no change in the number of cells required, area and clock cycles. In the shift register using D flip flop design involves 37 cells and delay of 3 clock cycles. The area required for this QCA design is 3,64 nm 2. Figure 6.3 Layout of SIPO Shift right register using JK flip flop Figure 6.32 Layout of SIPO Shift left register using JK flip flop Figure 6.33 Layout of SIPO Shift right register using D flip flop
26 4 Figure 6.34 Layout of SIPO Shift left register using D flip flop The performance analyses of the proposed shift registers are shown in Table 6.8. The performance analyses of those circuits are compared according to the complexity, area, and number of clock cycles. Table 6.8 Comparison of Shift Registers Shift Register Complexity Area Delay (clock cycles) SISO (D FF) 3 cells 565nm x 56nm 3 SIPO(D FF) 37 cells 565nm x 56nm 3 SISO/SIPO (JK FF) 28 cells 743nm x 36nm DESIGN OF COUNTERS IN QCA A counter is a sequential circuit consisting of set of flip flops connected in a suitable manner to count the sequence of the input pulses. The counters can be broadly classified into 3 types as follows:. Asynchronous and Synchronous counters 2. Single and multimode counters 3. Modulus counters An asynchronous or ripple counter can be constructed using minimum hardware. In an asynchronous counter, each flip-flop is triggered by
27 42 the output from the previous flip-flop which limits its speed of operation. The settling time in asynchronous counters, is the cumulative sum of the individual settling times of flip-flops. It is also called a serial counter. In synchronous counters, the speed limitation of ripple counter is overcome by applying clock pulses simultaneously to all the flip flops which leads to the settling time of the counter being equal to the propagation delay of a single flip-flop. Hence synchronous counters also are also called parallel counters. Single mode counters operate in single mode, i.e., it counts either in the Up mode or in the DOWN mode, whereas multimode counters operate in both UP and DOWN modes. Modulus counters are defined based on the number of states they are capable of counting. For example, a MOD counter has states. Counters are fundamental components of digital systems. Digital counters find wide application like pulse counting, frequency division, time measurement and control and timing operations. In this thesis, the 4-bit ring and Johnson counters are designed using D flip flops. The 3-bit up counter is designed using JK flip flops which counts upward from to 7. The performance analyses of those circuits are compared according to the complexity, area, and number of clock cycles Design of Ring Counter Ring counters are implemented using shift registers. It is essentially a circulating shift register connected so that the last flip-flop shifts its value into the first flip-flop. There is usually only a single circulating in the register, as long as clock pulses are applied.
28 43 Figure 6.35 Ring counter using D flip flop Table 6.9 Truth table of Ring counter In the diagram as shown in Figure 6.35, assuming a starting state of Q = and Q = Q 2 = Q 3 =. At the first pulse, the shift from Q to Q and the counter is in the state. The next pulse produces the state and the third,. At the fourth pulse, the at Q 3 is transferred back to Q, resulting in the state, which is the initial state. Subsequent pulses will cause the sequence to repeat, hence the name ring counter. This operation is shown in Table 6.9. The ring counter above functions as a MOD-4 counter since it has four distinct states and each flip-flop output waveform has a frequency equal to one-fourth of the clock frequency. A ring counter can be constructed for any MOD number. A MOD-N ring counter will require N flip-flops connected in the arrangement as the diagram above.
29 44 A ring counter requires more flip-flops than a binary counter for the same MOD number. For example, a MOD-8 ring counter requires 8 flip-flops while a MOD-8 binary counter only requires 3 (2 3 = 8). So if a ring counter is less efficient in the use of flip-flops than a binary counter, why do we still need ring counters? One main reason is because ring counters are much easier to decode. In fact, ring counters can be decoded without the use of logic gates. The decoding signal is obtained at the output of its corresponding flip-flop. Mehdi Askari et al (28) have designed ring counter using QCA cells. For the ring counter to operate properly, it must start with only one flip-flop in the state and all the others at. Since it is not possible to expect the counter to come up to this state when power is first applied to the circuit, it is necessary to preset the counter to the required starting state before the clock pulses are applied. One way to do this is to apply a pulse to the PRESET input of one of the flip-flops and the CLEAR inputs of all the others. This will place a single in the ring counter QCA Implementation of Ring Counter The flip flop is a significant part of the counters. In fact, a D-type flip flop can be constructed by a QCA binary wire with four clocking zones. An optimal design can be attained with attaching four optimal D flip flops as well as using an appropriate feedback being able to transmit output signal to input at every clock cycle. Figure 6.36 Layout of Ring counter using D flip flop
30 45 The 4-bit QCA ring counter design is shown in Figure Every different subsequent four clock zones is repeated alternatively and operates as a D-FF, i.e., the clocked binary wire operates as a D FF. It is necessary that D-FF3 output shift to D-FF after every clock cycle. So, feedback path only includes a complete clock cycle. There is a majority gate at the circuit input. Since one of its inputs has logic, majority gate is operated as a logic OR function. The set input, for the purpose of the initial turn on the circuit has been connected to the first flip-flop. The total number of cells required to implement a ring counter is 82, with an area of 84 nm 2. This is achieved by using cell minimization techniques Johnson Counter using DFF The Johnson counter, also known as the twisted-ring counter, is exactly the same as the ring counter except that the inverted output of the last flip-flop is connected to the input of the first flip-flop. This counter circulates a stream of ones followed by zeros around the ring. For example, a four bit shift counter with initial value of the counter of is followed by the repeating pattern,,,,,,,,. Figure 6.37 Johnson counter using D flip flop
31 46 Table 6. Truth table of Johnson counter Shift Pulse Q Q Q2 Q The Johnson counter works in the following way: Take the initial state of the counter to be. On the first clock pulse, the inverse of the last flip-flop will be fed into the first flip-flop, producing the state. On the second clock pulse, since the last flip-flop is still at level, another will be fed into the first flip-flop, giving the state. On the third clock pulse, the state is produced and on the fourth clock pulse, the state is produced. On the fifth clock pulse, the inverse of the last flip-flop, now a, will be shifted to the first flip-flop, giving the state. On the sixth and seventh clock pulse, using the same reasoning, we will get the state s and. On the ninth clock pulse, we will get the states, which is the initial state again. Hence, this Johnson counter has sixth distinct states:,,,,,, and, and the sequence is repeated so long as there is input pulse. Thus this is a MOD-8 Johnson counter. The MOD number of a Johnson counter is twice the number of flipflops. In the example above, four flip-flops were used to create the MOD-8 Johnson counter. So for a given MOD number, a Johnson counter requires only half the number of flip-flops needed for a ring counter. However, a Johnson counter requires decoding gates whereas a ring counter doesn't. As with the binary counter, one logic gate (AND gate) is required to decode each state, but with the Johnson counter, each gate requires only two inputs, regardless of the number of flip-flops in the counter. Note that we are
32 47 comparing with the binary counter using the speed up technique discussed above. The reason for this is that for each state, two of the N flip-flops used will be in a unique combination of states. Recirculating a single one, a ring counter divides the input clock by a factor equal to its number of stages whereas a Johnson counter divides by a factor equal to twice the number of stages. For example, a four stage Johnson counter divides by eight while a four stage Ring counter divides by four. Latching is done by using four clocking arrangement. Input is delivered to the output after one complete clock cycle delay. The four bit shift counter and the clocking of D flip flop is shown in Figure The truth table of Shift or Johnson Counter is shown given by Table 6.. A Johnson counters represent a middle ground between ring counters and binary counters. A Johnson counter requires fewer flip-flops than a ring counter but generally more than a binary counter; it has more decoding circuitry than a ring counter but less than a binary counter. Thus, it sometimes represents a logical choice for certain applications. Chakrabarty et al (2) have proposed QCA design of Johnson counter QCA implementation of Johnson Counter The 4-bit QCA implementation of Johnson counter design is shown in Figure It is constructed from Ring counter as Figure In this counter the inverted output of last flip flop is connected to the input of first flip flop. This is achieved by using QCA inverter which is indicated by a circle in the QCA layout as shown below.
33 48 Figure 6.38 Layout of Johnson counter using D flip flop Majority gate is used at the input for OR function for logic to initial turn on the circuit with the set input which is connected to the first flip flop. Once the set input is activated the circuit operates and the outputs are generated accordingly. In this work 93 cells are used, each of size 2nm x 2nm and area 984 nm 2 while the dot size is 5 nm, and effective area is 96, nm Design of Up Counter in QCA A synchronous binary counter counts from to 2 N -, where N is the number of bits/flip-flops in the counter. Each flip-flop is used to represent one bit. The flip-flop in the lowest-order position is complemented/toggled with every clock pulse and a flip-flop in any other position is complemented on the next clock pulse provided all the bits in the lower-order positions are equal to. Take for example A 4 A 3 A 2 A =. On the next count, A 4 A 3 A 2 A =. A, the lowest-order bit, is always complemented. A 2 is complemented because all the lower-order positions (A only in this case) are 's. A 3 is also complemented because all the lower-order positions, A 2 and A are 's. But A 4 is not complemented the lower-order positions, A 3 A 2 A =, do not give an all condition.
34 49 To implement a synchronous counter, we need a flip-flop for every bit and an AND gate for every bit except the first and the last bit. The diagram below shows the implementation of a 3-bit synchronous up-counter. Figure 6.39 Up counter using JK flip flops Table 6. Truth table of up counter A B C QCA implementation of Up Counter in QCA In conventional sequential circuits, any sequential circuit can be modelled as a single finite state machine (FSM) consisting of combinational logic and flip-flops. This model can also be used for QCA implementations of sequential logic. However, there exists an essential difference between
35 5 conventional and QCA sequential circuits. In a conventional sequential circuit, the clock signal controls only flip-flops. In a QCA sequential circuit, the clock signal controls not only flip-flops, but also the combinational logic. This difference brings many difficulties in QCA counter designs. Huang et al (27) have worked on the design of QCA counters. In this counter, the clock inputs of all the flip flops are connected together so that the input clock signal is applied simultaneously to each flip flop. Only the J and K inputs of FF is permanently connected to + while the J and K inputs of the other flip flops are driven by some combination of flip flop outputs. The J and K inputs of FF2 is connected with (A ) output of FF; the J and K inputs of FF3 is connected with AND operated output of A and B. Figure 6.4 Layout of Up counter using JK flip flops The 3-bit QCA up counter design is shown in Figure 6.4. It is constructed by using JK flip flop. It is a 3 bit counter hence it consists of 3 JK flip flops. Here the Majority gate is designed to perform AND operation of the output of flip flop and flip flop 2. This design uses 28 QCA cells and span over 2 clocking zones from input to output. The input-output delay of the circuit is therefore 3 clock cycles. The area required for this QCA design is 267,48 nm 2.
36 5 The performance analyses of proposed counters are shown in Table 6.2. The performance analyses of those circuits are compared according to the complexity, area, and number of clock cycles. The proposed designs are compared with existing majority gate method. Table 6.2 Comparison of counters Previous Structures Proposed structures Delay Circuit (clock Complexity Area Complexity Area Name cycles) Ring Counter 86 cells 75nm x 2nm 82 cells 7nm x 2nm 5 Johnson Counter 99 cells 86nm x 2nm 93 cells 8nm x 2nm 5 Up Counter 22 cells 8nm x 34nm 28 cells 743nm x 34nm 4 The length of the structure increases as the number of cells increases. But the width of the structure remains the same. Therefore the difference in the number of cells is not proportional to the area. 6.5 RESULTS AND DISCUSSION The design of QCA sequential circuits is quite different from the design of traditional CMOS sequential circuits because of their different clocking schemes. The four types of flip flops have implemented and simulated here. The simulated results of all these flip flops are shown from Figure 6.4to Figure The simulation results obtained with QCADesigner ver.2..3 verifies the functionality of the proposed flip flops. The delay for the each flip flop is clock cycle. Therefore, each circuit has four clocking zones. Initially clock is used to get the inputs. Clock is used to route inputs for
37 52 majority gate logic, clock 2 is used for finding majority logic and clock 3 is used to compute output. The output is available at clock again. Figure 6.4 Simulation Result of SR flip flop Figure 6.42 Simulation Result of D flip flop Figure 6.43 Simulation Result of JK flip flop
38 53 Figure 6.44 Simulation Result of T flip flop The simulated results of 3 bit serial in serial out and serial in parallel out shift registers are shown in Figure 6.45 and Figure The shift registers are constructed by 3 flip flops. Hence 3clock cycles or 2 clock zones are required to transmit the data from input to the output. In SISO shift register, at each clock cycle single bit is shifted to next flip flop either right or left direction as shown in Figure But the SIPO shift register, all the bits are available simultaneously as shown in Figure Figure 6.45 Simulation Result of SISO shift register
39 54 Figure 6.46 Simulation Result of SIPO shift register The simulated result of 4-bit QCA ring counter is shown in Figure The counter is constructed by D flip flops. The D flip flop is designed by a QCA binary wire as shown in Figure It is 4 bit counter hence it consists of 4 D flip flops. Therefore, 4 clock cycles are needed for the design of 4 flip flops and clock cycle is used for feedback path. Hence 5 clock cycles are required for the proper operation of the circuit. The set input, is used to turn on the circuit initially which is connected to the first flip flop. The Set input waveform and circuit outputs are shown in Figure In this figure, the rotation of bit is indicated properly. With QCADesigner ver.2..3, the circuit functionality is verified and this design has 82 cells and an area of approximately 84 nm 2. Figure 6.47 Simulation Result of Ring counter
40 55 The simulated result of 4-bit QCA Johnson counter is shown in Figure Similar to ring counter it is also constructed by D flip flops. The D flip flop is designed by a QCA binary wire as shown in Figure The delay for the circuit is 5 clock cycles which is required to produce the output. Once the set input is activated the circuit operates and the outputs are generated accordingly. Figure 6.48 Simulation Result of Johnson counter The proposed QCA up counter has been designed and simulated using the QCADesigner tool. This tool allows users to do a custom layout and then verify QCA circuit functionality by simulations. The simulated result of 3-bit QCA up counter is shown in Figure This counter is constructed by JK flip flops as shown in Figure 6.4. The delay for the circuit is 4 clock cycles which is required to produce the output. Here the 3 clock cycles are used to design the 3 JK flip flops and clock cycle is used for combinational logic (MG) design.
41 56 Figure 6.49 Simulation Result of Up counter 6.5 CONCLUSION In this chapter the different flip flops have been designed using majority gates. The layouts and functionality checks were done using QCADesigner and the designs are compared according to the complexity, area and number of clock cycles. The operations of these circuits have been verified according to the truth table. The proposed layouts are significantly smaller than the same circuits using standard CMOS technology as well as the existing circuits in QCA. The different types of shift registers and counters have been designed and analyzed with the help of proposed D and JK flip flops. It is shown that clocking requirements, number of zones, as well as the underlying CMOS circuitry complexity are significantly reduced compared with previous QCA circuits.