Adiabatic charging register circuit

Details Adiabatic charging register circuit Adiabatic charging register circuit

2001-06-04

2004-09-07

Callahan, Timothy P. (Department: 2816)

Miscellaneous active electrical nonlinear devices, circuits, and

Signal converting, shaping, or generating

Particular stable state circuit

C327S218000

Reexamination Certificate

active

06788121

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an adiabatic charging register circuit and, in particular, relates to such a circuit which reduces power consumption associated with a clock pulse.
2. Description of the Related Art
Conventionally, an LSI circuit includes a large number of register circuits, fifty thousand or more circuits. Each register circuit comprises a D-flip flop (D-FF) or D-latch circuit. A register circuit having a D-FF or D-latch is exemplified here.
First, a register circuit having a D-FF is described.
A D-FF has a pair of D-latch circuits.
FIG. 17
shows an example of a D-latch circuit (page 677, “Structure and Design of a Computer”, by David A. Patterson and John L. Hennessy, published by NikkeiBP). A D-latch circuit
70
in
FIG. 17
has a pair of NOR circuits
71
and
72
which are cross-connected to each other constituting a RS-FF (reset-set-flip-flop), and a pair of AND circuits
73
and
74
. It has data input terminals D, DN in differential form, a clock input terminal CK, and data output terminals Q, QN in differential form. Each of NOR circuits
71
and
72
operates as an inverter when one of the inputs of the same is in low state, and therefore, a D-latch circuit
70
in
FIG. 17
operates as follows.
(1) A pair of NOR circuits
71
,
72
keep a previous state when a clock input terminal CK is in low state.
(2) A value of a data input terminal D is stored in a pair of NOR circuits
71
,
72
when a clock input terminal CK is in high state.
A D-FF circuit
80
is constituted by using a pair of D-latch circuits
70
as shown in
FIG. 18
, in which a first stage D-latch circuit
70
A receives a clock input CK as it is, and a second stage D-latch circuit
70
B receives a clock signal which is inversion by 180° of the clock input CK by an inverter
90
. A D-FF circuit operates as follows.
(1) When a clock input CK becomes a high state, the first stage D-latch circuit
70
A opens to accept a data D at an input terminal,
(2) When a clock input CK becomes a low state, the second stage D-latch circuit
70
B opens and an input terminal D accepts an output O
1
on an output terminal Q of the first stage D-latch
70
A, as an input signal D.
FIG. 19
shows operational wave forms of a data D, a clock input CK, an output O
1
of the first stage D-latch circuit
70
A, and an output O
2
of the second stage D-latch circuit
70
B. As shown in
FIG. 19
, the output O
2
is switched by an input data when a clock input CK is switched from high state to low state, thus, it is an edge trigger type circuit.
Conventionally, a clock signal CK is generated by using an inverter having a CMOS circuit which has a p-channel MOSFET and an n-channel MOSFET connected in series with each other, and has rectangular wave form. A load coupled with an output of a clock signal generator is charged to power supply voltage VDD through p-channel MOSFET of an inverter when an output signal is in high state, and is discharged to ground through n-channel MOSFET of an inverter when an output signal is in a low state. Therefore, the power consumption P by a clock signal is P=CV
2
f, where f is the clock frequency, V is the power supply voltage, and C is the sum of capacitance of wires and gate capacitance which accept a clock signal.
The capacitance of wires is recently large because of an increase of semiconductor chip area of an integrated circuit reflecting a large scale integrated circuit, and therefore, power consumption by charge/discharge of a clock signal occupies almost 50% of the total power consumption of a semiconductor chip (page 90, Technical Report of Low Power LSI, Nikkei Micro-device, NikkeiBP).
Further, a large number of register circuits are used for a pipeline processing in an LSI for processing a moving image, and a RISC processor. In those devices, it is also known that power consumption by a clock system is almost the same as that by a logic system (page 8, Low power and high speed LSI technology, Realize Co.). That relation is independent from operation speed, but depends upon the ratio occupied by a register circuit in an LSI.
SUMMARY OF THE INVENTION
It is an object, therefore, of the present invention to provide a new and improved register circuit by overcoming the disadvantages and limitations of a prior register circuit.
It is also an object of the present invention to provide a register circuit which consumes less power in a clock system.
It is further an object of the present invention to provide a register circuit in which no short-circuit current from a power source to ground directly flows.
The above and other objects are attained by an adiabatic register circuit comprising; a plurality of n-channel MOSFET transistors and a plurality of p-channel MOSFET transistors, accepting an input data, and a clock signal, and providing an output data; said clock signal being a power clock signal having a gradually rising and gradually falling waveform generated by using a charge recycle power source in which power supplied to a load is at least partially collected and returned to said charge recycle power source; and following inequality is satisfied;
|
V
TN
|+|V
TP
|≧VDD
where V
TN
is threshold of said n-channel MOSFET transistor, V
TP
is threshold of said p-channel MOSFET, and VDD is output voltage of said charge recycle power source.
Preferably, said register circuit comprises a pair of D-latch circuits with an input of a second D-latch circuit coupled with an output of a first D-latch circuit, a first D-latch circuit accepts a first power clock signal, and a second D-latch circuit accepts a second power clock signal which is different by 180° of the first power clock signal.
Preferably, said D-latch circuit comprises a pair of NOR circuits with one of the inputs of each NOR circuit being coupled with an output of the other NOR circuit, and a pair of AND circuits each accepting an input data in differential form and a power clock signal, and providing an output to the other input of each of said NOR circuit.
Preferably, said register circuit includes a combination logic circuit between said pair of D-latch circuits.
Preferably, said D-latch circuit comprises a memory element having a first inverter providing an output of the D-latch circuit, a second inverter with an input coupled with an output of said first inverter, and a first transmission gate connecting an output of the second inverter to an input of the first inverter, and a second transmission gate inserted between an input terminal and an input of said first inverter.


REFERENCES:
patent: 5473526 (1995-12-01), Svensson et al.
patent: 5521538 (1996-05-01), Dickinson
patent: 5900758 (1999-05-01), Kanno
patent: 5994935 (1999-11-01), Ueda et al.
patent: 6046648 (2000-04-01), Nakamiya et al.
patent: 6313673 (2001-11-01), Watanabe
patent: 6323709 (2001-11-01), Kulkarni et al.
patent: 8-335873 (1996-12-01), None
patent: 9-74347 (1997-03-01), None
patent: 10-190442 (1998-07-01), None
patent: 10-308662 (1998-11-01), None
Athas, William C., “Energy-Recovery CMOS”, Low power Design Methodologies, J. M. Rabaey and M. Pedram (Kluwer Academic Publishers, 1996), Chapter 4, pp. 65-72.
Athas, William C. et al., “A Low-Power Microprocessor Based on Resonant Energy”, IEEE Journal of Solid-State Circuits, vol. 32, No. 11, Nov. 1997, pp. 1693-1701.
“Low Power and High Speed LSI Technology”, Realize Co., along with partial translation of relevant parts, Jan. 31, 1998.
Patterson et al., Structure and Design of a Computer, by David A. Patterson, and John L. Hennessy, published by NikkeiBP, Sep. 22, 1997, p. 677.
Technical Report of Low Power LSI, Nikkei Micro-device, NikkeiBP, 1994, p. 90.
“A Low Power Multiplier Using Adiabatic Charging Binary Decision Diagram Circuit”,Extended Abstracts of the 1999 International Conference on Solid State Devices and Materials, Tokyo, 1999, pp. 444-445.
Svensson et al., “Low Power Circuit Techniques”,Low Power Design Methodologies, Kluwer Academic Publishers, 1996, Chapter 3, pp. 37-52.

Affiliated with

Also associated with

Rating

Adiabatic charging register circuit does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Adiabatic charging register circuit, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Adiabatic charging register circuit will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3266112