Semiconductor memory integrated circuit

Details Semiconductor memory integrated circuit Semiconductor memory integrated circuit

2001-05-25

2002-12-31

Nguyen, Tan T. (Department: 2818)

Static information storage and retrieval

Addressing

Particular decoder or driver circuit

Reexamination Certificate

active

06501702

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention generally relates to clock synchronization type semiconductor memory integrated circuits and, more particularly, to an improvement of an address decoding circuit.
2. Related Background Art
FIG. 14
shows a block diagram of a prior art clock synchronization type semiconductor memory integrated circuit. An external clock signal CLK is inputted to a clock buffer
4
, in which internal clock signals CK
1
and CK
2
are generated. Address signals are inputted to an address buffer
8
in sync with the clock signals CK
1
, among which a row address signal is sequentially decoded by a predecoder
6
and a row decoder
7
to select a word line WL of a memory cell array
1
.
The internal clock signal CK
2
from the clock buffer
4
and a timing pulse PULSE generated at a pulse generating circuit
5
based on the internal clock signal CK
1
are sent to the predecoder
6
. The timing pulse PULSE is used for adjusting a transfer timing of a predecoded output PD
1
to the row decoder
7
.
A column address is decoded by a column decoder
9
. Based on the decoded output from the column decoder
9
, a column gate
3
is selectively driven to select a bit line of the memory cell array
1
. Data of a memory cell MC determined by the selected word line WL and the selected bit line BL is read out to the bit line BL, sensed and amplified by a sense amplifier
2
, and outputted via a data buffer
10
.
In a large-scale semiconductor memory circuit, the memory cell array
1
is divided into a plurality of memory sections
1
-
1
to
1
-n, as shown in
FIG. 15. A
main word line MWL is shared by the memory sections
1
-
1
to
1
-n, and a plurality of word lines WL, each corresponding to one of the memory sections, are selectively connected to the main word line MWL in accordance with section selection signals SSL
1
to SSLn.
FIG. 16
shows an example of a specific configuration of the row address decoding system (word line selecting system) shown in FIG.
14
. The address buffer
8
includes a latch circuit (master latch), having inverters INV
4
and INV
5
, for temporally holding address data inputted in sync with the clock signal CK
1
. In order to carry out a grouping selection of a plurality of main word lines MWL, the predecoder
6
includes a predecoding section
6
a
for selecting one line from a bundle of main word lines MWL, and a predecoding section
6
b
for selecting a bundle of main word lines MWL.
The predecoding sections
6
a
and
6
b
have NAND gates NAND
1
and NAND
3
, respectively, for decoding an internal address A
0
sent from the address buffer
8
in accordance with the clock signal CK
2
. Further, the predecoding sections
6
a
and
6
b
have latch circuits (slave latches) formed of inverters (INV
9
, INV
10
) and (INV
13
, INV
14
), respectively. The decoded output latched by the predecoding section
6
a
is taken from a NAND gate NAND
2
and sent to the row decoder
7
as an output PD
1
. The latch data of the other predecoding section
6
b
is sent to the row decoder
7
as an output PD
2
via inverters INV
15
and INV
16
without being adjusted by the timing pulse.
The pulse generating circuit
5
generates a timing pulse PULSE in consideration of the number of logical circuits (delays) provided before the predecoder
6
decodes an address externally received. Specifically, in the example shown in
FIG. 16
, odd-number inverters INV
24
to INV
26
and a NAND gate NAND
5
form an edge detecting circuit
51
for detecting a rising edge of the internal clock signal CK
1
, and even-number inverters INV
28
to INV
31
and a NOR gate NOR
1
form a pulse width setting circuit
52
for determining a pulse width. Therefore, the pulse generating circuit
5
generates, in sync with a rising of the clock signal CK
1
, a timing pulse PULSE having a pulse width determined by the delays of the inverters INV
28
to INV
31
.
The row decoder
7
includes a NAND gate NAND
4
for further decoding the predecoded output PD
2
and transfer gate transistors M
1
and M
2
, which are controlled by the output of the NAND
4
and transfer the predecoded output PD
1
to the main word line MWL. In more detail, in the row decoder, when the prededecoded output PD
2
is inputted (all 1's), the output of the NAND gate NAND
4
becomes “L”, thereby turning on the transfer gate transistors M
1
and M
2
to supply the predecoded output PD
1
to the main word line MWL.
FIG. 17
shows the timings in the above described row address decoder from the input of the external clock signal CLK to the activating of a word line. A timing pulse PULSE is generated so as to have a predetermined pulse width in which the predecoded outputs PD
1
and PD
2
are accomplished. Thus, the predecoded outputs PD
1
and PD
2
are further decoded by the row decoder
7
, and the selection of a main word line MWL and a selection of a word line ML are carried out. In addition, in the row decoder
7
, three inverters INV
18
to INV
20
inserted after the transfer gate transistors M
1
and M
2
are for the purpose of waveform formation, because an “H” level attenuation occurs in the transfer gate transistors M
1
and M
2
, and the signal level is not fully compensated by only a single inverter. After a main word line MWL is selected, a word line WL is selected by a NOR gate NOR
2
activated by a section selection signal SSL.
In the prior art clock synchronization type address decoder shown in
FIG. 16
, it is preferable to reduce the delay determined by the number of the logical circuits in the address buffer
8
and the predecoder
6
, and to speed up the rising of a timing pulse PULSE generated by the pulse generating circuit
5
as much as possible. Since these sections are already formed of a minimum number of logical circuits, it is difficult to remarkably speed up the operations of these sections unless it is possible to reduce the delay of each logical circuit.
In the circuit arrangement shown in
FIG. 16
, it is rather the number of logical circuits needed to be used after the generation of the timing pulse PULSE until the selection of a word line WL that is preventing the speeding up of the operations. In
FIG. 16
, if the transfer gate transistors (M
1
, M
2
) are assumed to be a single logical circuit, the number of the logical circuits after the generation of the timing pulse PULSE until the selection of a word line WL is seven (NAND
2
, INV
11
, (M
1
, M
2
), INV
18
to INV
20
, and NOR
2
). Due to the existence of these logical circuits, a considerable delay is caused before a word line is activated. In
FIG. 17
, a delay after the rising of the timing pulse PULSE until the activation of the selected word line WL is shown as &tgr;
2
.
Further, as the capacities of semiconductor memories increase, if the number of bits in a row system is doubled, the number of row decoder is also doubled, thereby doubling the load per one predecoder. Accordingly, the delays caused by the predecoded outputs PD
1
and PD
2
are increased. In order to reduce such delays, the size of output buffer in the predecoder, i.e., the size of the inverters INV
11
and INV
16
in
FIG. 16
, maybe increased. However, there is a limit to the improvement of input waveform of a row decoder, which causes a large load. Further, there may arise another problem in that the increasing of the size of output buffer may increase power consumption.
SUMMARY OF THE INVENTION
Given the above mentioned circumstances, an object of the present invention is to provide a semiconductor memory integrated circuit in which the number of logical circuits needed to be used during a period from the input of a clock signal to the activation of a word line is reduced to speed up operations.
A semiconductor memory integrated circuit according to the present invention comprises: a memory cell array; a clock buffer for receiving a clock signal to generate an internal clock signal; an address buffer for receiving an address signal in accordance with said internal clock signal generated by said clock buffer; a

Affiliated with

Also associated with

Rating

Semiconductor memory integrated 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 Semiconductor memory integrated circuit, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Semiconductor memory integrated circuit will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2939000