Professional Documents
Culture Documents
Implementation of Circuit in Different Adiabatic Logic: Abstract-Adiabatic Circuits Are Widely Employed in Low Power
Implementation of Circuit in Different Adiabatic Logic: Abstract-Adiabatic Circuits Are Widely Employed in Low Power
IMPLEMENTATION OF CIRCUIT IN
DIFFERENT ADIABATIC LOGIC
3
MOUMITA BHOWMIK,
4
ARPAN CHAUDHURI MAMIA SAHA SAMBHU NATH PRADHAN
5
CAPGEMINI RELIANCE JIO INFOCOMM SUBHRAJIT DAS
5
HYDERABAD LIMITED PUNJAB NATIONAL BANK
INDIA MUMBAI ECE DEPARTMENT
arpanchaudhuri18@gmail.com INDIA NIT AGARTALA
mamiasaha20@gmail.com AGARTALA, INDIA
moumitabk23@gmail.com,
sambhu.pradhan@gmail.com
subhrajit54@gmail.com
Abstract- Adiabatic circuits are widely employed in Low power Since, the AC supplies controls the working of the circuits,
VLSI circuit to achieve power efficient system at the cost of they are also called power clocks.
reduced performance. The power saving of adiabatic circuit Over the years different low power adiabatic logic circuits
can reach more than 90% compared to conventional static are proposed [1-4]. Though, CMOS technology provides
CMOS logic for extreme low frequency applications. The circuits with very low static power dissipation, during the
power efficiency of 4:1 multiplexer, designed in different switching operation currents are generated, due to the discharge
adiabatic logic families is presented in this paper. Power of load capacitances that cause a power dissipation which
saving more than 50% is achievable beyond 1000MHz also. increases with the clock frequency [5]. The adiabatic technique
prevents such losses. The charge does not flow from the load
Keywords—Low-power,adiabatic logic,multiplexer, CMOS, capacitance to ground, but it flows back to a trapezoidal or
Positive feed back adiabatic logic, Efficient charge recovery sinusoidal supply voltage and can be reused [6]. Power losses
logic,2N2P2N. due to the resistance of the switches needed for the logic
operation still occur. In order to keep these losses small, the
clock frequency has to be much lower than the technological
I. INTRODUCTION limit.
With the development of VLSI technology power Multimode operations of adiabatic circuits are presented
dissipation is increasing dramatically. Low power has become in [7]. In this paper a simple combinational circuit -4:1
one of the crucial design constraint, especially for portable and multiplexer is implemented using different configuration of
battery operated systems. Compared with the conventional low adiabatic logics and the power saving in different frequency
power approaches, power dissipation can be significantly are compared to see the effeteness of different adiabatic
reduced by using the adiabatic computation. logic families with respect to low power consumption. The
The term adiabatic comes from thermodynamics, used to rest of the paper is organized as follows. Section II starts
describe a process in which there is no exchange of heat with with the discussion on working principal and power
the environment. The adiabatic logic structure dramatically dissipation of conventional CMOS and adiabatic circuit.
reduces the power dissipation. The adiabatic switching Section II ends with the description of different adiabatic
technique can achieve very low power dissipation, but at the logic families. Simulation results of adiabatic multiplexer
expense of circuit complexity. Adiabatic logic offers a way to (mux) are presented in section III. The paper ends with the
reuse the energy stored in the load capacitors rather than the conclusion given in section IV.
traditional way of discharging the load capacitors to the ground
and wasting this energy. By properly mixing the ideas derived II. CONVENTIONAL AND ADIABATIC LOGIC
for adiabatic and static CMOS circuits one can achieve low
power dissipation in the circuit. Adiabatic logic circuits utilize A. CONVENTIONAL CMOS LOGIC
AC voltage supplies (power-clocks) to recycle the energy of The conventional switching can be understood by using a
circuit nodes. During recovery phase, the energy of the circuit simple CMOS inverter. The CMOS inverter can be
nodes is recovered to the power source instead of being considered to consist of a pull-up and pull-down networks
dissipated as heat. In the adiabatic circuits, circuit nodes are connected to a load (or internal) capacitance C. The pull-up
charged and discharged by AC voltage supplies, thus their and pull-down networks are actually MOS transistors in
output signals are clocked AC signals (adiabatic signals). series with the same load C. The capacitance in this case
models the fan-out of the output signal.
354
Figure 4. An Adiabatic logic gate
C. STEPS OF ADIABATIC CIRCUIT DESIGN Figure 5. General Schematic (left) and timing of NOT gate
General rules or steps to be followed for adiabatic (right) for ECRL
circuit design are as follows:
E.POSITIVE FEEDBACK ADIABATIC LOGIC (PFAL)
x Replace each of the PMOS and NMOS devices in The structure of PFAL logic is shown in Fig. 6. Two n-
the pull-up and pull-down networks with T-gates. trees realize the logic functions. This logic family also
generates both positive and negative outputs [4,5].
x Use expanded pull-up network to drive the true
output.
x Replace DC V dd by a pulsed power supply (Vpwr) Figure 6. General Schematic for PFAL family
with varying voltage to allow adiabatic operation.
Today’s adiabatic circuit employs MOSFET in place of T- The two major differences with respect to ECRL are that
gates. the latch is made by two pMOSFETs and two nMOSFETs,
In this work, out of the many adiabatic logic families rather than by only two pMOSFETs as in ECRL, and that
following three logic families are chosen. the functional blocks are in parallel with the transmission
pMOSFETs. In the same way, energy recycling is there.
x Efficient Charge Recovery Logic (ECRL)
F.2N-2N2P LOGIC
x Positive Feedback Adiabatic Logic (PFAL) This adiabatic logic family is derived from ECRL in order
to reduce the coupling effect. Fig.7 shows the general
schematic of ECRL logic [3]. The primary advantage of 2N-
x 2N-2N2P
2N2P over ECRL is that the cross-coupled nMOSFETs
switches result in non-floating outputs for large part of the
D. EFFICIENT CHARGE RECOVERY LOGIC (ECRL)
recovery phase.
Efficient Charge Recovery Logic (ECRL) proposed by
This family is a variation of the 2N-2N2P family and
Moon and Jeong , shown in Figure 5, uses cross-coupled
consists of 2 new cross-coupled nMOS transistors added in
PMOS transistors [2]. It has the structure similar to Cascade
parallel to the 2 nMOS transistors. The timing and operation
Voltage Switch Logic (CVSL) with differential signaling. .It
of this family is identical to the 2N-2N2P p family and the
consists of two cross-coupled transistors M1 and M2 and
buffer/inverter is shown in Fig. 7.
two NMOS transistors.
A 4:1 multiplexer is designed and simulated to get the
Fig.5 shows the general schematic for ECRL and the
power dissipation result using Conventional CMOS logic,
waveforms of the supply clock as well as I/O signals for a
2N-2N2P logic, ECRL logic and PFAL logic.
NOT gate. In order to recover and to reuse the supplied
Also, we are going to compare the power results obtained
energy, an ac power supply is used for ECRL gates.
from them in different frequencies to ensure the benefits of
355
adiabatic logic designing over conventional CMOS 30ps
designing. A 4:1 Multiplexers has four inputs. A, B, C, D ܸܽ pulse of T=100ps
and two select lines ܵ0 , ܵ1 .The output is Y = (A. ܵഥ0 ܵഥ1 Ĭġ ܸܽ :Square
pulse Rise Time: 30ps, Fall Time:
(B. ܵഥ0 ܵ1 ) + (C. ܵ0 ܵഥ1 ġĩŅįܵ0 ܵ1 ). 30ps
Input
ܸܾ :Square ܸܾ pulse of T=400ps
pulse Rise Time: 30ps, Fall Time:
30ps
Input ܸܿ :Square ܸܿ pulse of T=600ps
pulse Rise Time: 30ps, Fall Time:
30ps
ܸ݀ :Square ܸ݀ pulse of T=500ps
pulse Rise Time: 30ps, Fall Time:
30ps
Selection ܵ0 V pulse of T=900ps
lines Rise Time: 30ps, Fall Time:
30ps
ܵ1 V pulse of T=800ps
Rise Time: 30ps, Fall Time:
30ps
Figure 7. General Schematic for 2N-2N2P logic family
356
V. CONCLUSION adiabatic logic may well be applied in lower power
This paper reviews the basics of adiabatic computation and computation within this frequency. However, new adiabatic
the most well-known adiabatic logic families are described. design at high frequency may be targeted so that adiabatic
Out of several logic families, positive feedback adiabatic circuit may be used in many high frequency applications
logic (PFAL) gives the better results in terms of power also.
dissipation. Results upto 1 GHz has been shown and
357
Figure 11. Schematic of PFAL 4:1 MUX
358
Graphical Comparison Between CMOS logic 2N- Graphical Comparison Between CMOS logic 2N-
2N2P,ECRL & PFAL logic 2N2P,ECRL & PFAL logic
Power (mW)
100
80
Power ( mW)
80
60
60
40
40
20 20
0 0
0.1
1
10
100
200
500
700
800
900
1000
0.1
1
10
100
200
300
400
500
600
700
800
900
1000
Figure 12. Power plot over frequency. Figure 13. Power plot as bar chart over frequency
359