Patent Application
20230308099
The invention relates to a buffer circuit, and more particularly to a high-speed buffer circuit.
A buffer circuit is one of the most commonly used digital circuits in chip and processor design. A buffer operates to receive an input signal and output an output signal that keeps the characteristics carried by the input signal but has a better driving ability. In high-speed applications such as central processing units (CPUs), graphics processing units (GPUs), and accelerated processing units (APUs) the amount of signal delay is an important indicator for evaluating performance. Thus, buffer circuits that operate at high speeds are required to enhance the performance of high-speed applications.
An exemplary embodiment of a buffer circuit may pre-charge or pre-discharge an output node in response to an input, so that an output signal generated at the output node may rapidly reach a voltage level in response to the input signal.
An exemplary embodiment of a buffer circuit is provided. The buffer circuit receives an input signal at an input node and outputs an output signal at an output node. The buffer circuit includes a first inverter, a second inverter, a first switch, and a second switch. The first inverter is coupled to the input node to receive the input signal. The first inverter inverts the input signal to generate an inverted input signal. The second inverter has an input terminal receiving the inverted input signal and an output terminal coupled to the output node. The first switch is coupled between a first voltage source terminal and the output node and controlled by the input signal. The second switch is coupled between the output node and a second voltage source terminal and controlled by the input signal. A first voltage is provided to the first voltage source terminal. A second voltage is provided to the second voltage source terminal. In response to the input signal switching to a first level from a second level, the first switch is turned on to pre-charge the output node toward the first voltage. In response to the input signal transiting to the second level from the first level, the second switch is turned on to pre-discharge the output node toward the second voltage.
An exemplary embodiment of a semiconductor structure of a buffer circuit is provided. The semiconductor structure comprises a substrate, a P-type diffusion region, a N-type diffusion region, a first polysilicon region, a first conductive segment, and a second conductive segment. The P-type diffusion region is formed on the substrate. The P-type diffusion region extends parallel to a first direction. The N-type diffusion region is formed on the substrate and spaced apart from the P-type diffusion region. The N-type diffusion region extends parallel to the first direction. The first polysilicon region is formed above the P-type and N-type diffusion regions. The first polysilicon region extends parallel to a second direction. The first direction is perpendicular to the second direction. The first conductive segment extends parallel to the first direction. The second conductive segment is spaced apart from the first conductive segment. The second conductive segment extends parallel to the first direction. The P-type diffusion region comprises a first P-type diffusion portion disposed on a first side of the first polysilicon region, and the first P-type diffusion portion is electrically connected to the second conductive segment. The N-type diffusion region comprises a first N-type diffusion portion disposed on the first side of the first poly region, and the first N-type diffusion portion is electrically connected to the first conductive segment.
A detailed description is given in the following embodiments with reference to the accompanying drawings.
The invention can be more fully understood by reading the subsequent detailed description and examples with references made to the accompanying drawings, wherein:
The following description is of the best-contemplated mode of carrying out the invention. This description is made for the purpose of illustrating the general principles of the invention and should not be taken in a limiting sense. The scope of the invention is best determined by reference to the appended claims.
Referring to
The switch 12 is coupled between the voltage source terminal VS10 and the output node N11 and further coupled to the input node N10 to receive the input signal S10. The switch 12 is controlled by the input signal S10, in detail, the turned-on/off state of the switch 12 is determined according to the input signal S10. The switch 13 is coupled between the output node N11 and the voltage source terminal VS10 and further coupled to the input node N10 to receive the input signal S10. The switch 13 is controlled by the input signal S10, in detail, the turned-on/off state of the switch 13 is determined according to the input signal S10.
In the following paragraphs, the high-speed operation of the buffer circuit 1 is described.
Based on the operation of the buffer circuit 1, the output signal S11 generated at the output node N11 keeps the characteristics carried by the input signal S10, such as the phase, polarity, and/or voltage level of the input signal S10. For example, when the input signal S10 is at a relative high voltage level, the output signal S11 is at a relative high voltage level; when the input signal S10 is at a relative low voltage level, the output signal S11 is at a relative low voltage level. Accordingly, the output signal S11 varies with the input signal S10.
The inverter 10 receives the input signal S10 through the input terminal TI10. In cases in which the input signal S10 switches to the relative high voltage level from the relative low voltage level, the inverter 10 receives the input signal S10 having the relative high voltage level and then inverts the input signal S10 to generate an inverted input signal S12 having a relative low voltage level at the output terminal TO10. The switches 12 and 13 also receive the input signal S10. In the embodiment, according to the relative high voltage level of the input signal S10, the switch 12 is turned on while the switch 13 is turned off. At this time, the output node N11 is charged toward the voltage VDD through the turned-on switch 12. Since the switch 13 is turned off, a discharge path between the output node N11 and the voltage source terminal VS11 is cut off, so that the output node N11 can be charged toward the voltage VDD stably and continuously.
The inverter 11 receives the inverted input signal S12 having the relative low voltage level through the input terminal TI11 and then inverts the inverted input signal S12 to generate the output signal S11 at the output terminal TO11 that is coupled to the output node N11. As described above, since the output node N11 has been charged toward the voltage VDD, the voltage level of the output signal S11 can reach the relative high voltage level in a short time. In the embodiment, the relative high voltage level of the output signal S11 is equal to the level of the voltage VDD.
In cases in which the input signal S10 switches to the relative low voltage level from the relative high voltage level, the inverter 10 receives the input signal S10 having the relative low voltage level and then inverts the input signal S10 to generate the inverted input signal S12 having a relative high voltage level at the output terminal TO10. The switches 12 and 13 also receive the input signal S10. In the embodiment, according to the relative low voltage level of the input signal S10, the switch 12 is turned off while the switch 13 is turned on. At this time, the output node N11 is discharged toward the voltage VSS through the turned-on switch 13. Since the switch 12 is turned off, a charge path between the voltage source terminal VS10 and the output node N11 is cut off, so that the output node N11 can be discharged toward the voltage VSS stably and continuously.
The inverter 11 receives the inverted input signal S12 having the relative high voltage level through the input terminal TI11 and then inverts the inverted input signal S12 to generate the output signal S11 at the output terminal TO11 that is coupled to the output node N11. As described above, since the output node N11 has been discharged toward the voltage VSS, the voltage level of the output signal S11 can reach the relative low voltage level in a short time. In the embodiment, the relative low voltage level of the output signal S11 is equal to the level of the voltage VSS.
According to the embodiment of
For a convention buffer circuit comprising two inverter stages coupled in series without the switches 12 and 13 proposed by the present embodiment, if there is a heavy load coupled to the output terminal of the buffer circuit, the output signal generated by the buffer circuit may take a long time to reach a predetermined voltage level in response to the input signal of the buffer circuit, which increases the delay between the output signal and the input signal. The increases delay causes low performance of the circuitry using the buffer circuit.
On the contrary, the buffer circuit 1 provided by the embodiment can pre-charge or pre-discharge the output signal S11 in response to the voltage level of the input signal S10. Therefore, even if a heavy load is coupled to the output node N11 of the buffer circuit 1, the output signal S11 can still rapidly reach a predetermined voltage level in response to the switching of the voltage level of the input signal S10, thereby achieving the high-speed operation of the buffer circuit.
The difference between the buffer circuit 2 of
In cases in which the input signal S10 switches to the relative high voltage level from the relative low voltage level, according to the relative high voltage level of the input signal S10, the N-type transistor 22 is turned on while the P-type transistor 23 is turned off. At this time, the output node N11 is charged toward the voltage VDD through the turned-on N-type transistor 22. Since the P-type transistor 23 is turned off, a discharge path between the output node N11 and the voltage source terminal VS11 is cut off, so that the output node N11 can be charged toward the voltage VDD stably and continuously.
In cases in which the input signal S10 switches to the relative low voltage level from the relative high voltage level, according to the relative low voltage level of the input signal S10, the N-type transistor 22 is turned off while the P-type transistor 23 is turned on. At this time, the output node N11 is discharged toward the voltage VSS through the turned-on P-type transistor 23. Since the N-type transistor 22 is turned off, a charge path between the voltage source terminal VS10 and the output node N11 is cut off, so that the output node N11 can be discharged toward the voltage VSS stably and continuously.
According to the embodiment of
The difference between the buffer circuit 3 of
The gate electrode of the N-type transistor 30B is coupled to the output terminal TO10 of the inverter 10 to receive the inverted input signal S12, and the first and second drain/source electrodes thereof are coupled to the output node N11 and the voltage source terminal VS11 respectively. As shown in
In cases in which the input signal S10 switches to the relative high voltage level from the relative low voltage level, the P-type transistor 30A is turned on according to the inverted input signal S12 having the relative low voltage level. At the same time, the N-type transistor 22 is also turned on according to the input signal S10 having the relative high voltage level. Moreover, the transistors 23 and 30B are turned off. Due to the turned-on states of the transistors 12 and 30A, the transmission gate 31A passes the voltage VDD to the output node N11, thereby charging the output node N11 toward the voltage VDD.
In cases in which the input signal S10 switches to the relative low voltage level from the relative high voltage level, the N-type transistor 30B is turned on according to the inverted input signal S12 having the relative high voltage level. At the same time, the P-type transistor 23 is also turned on according to the input signal S10 having the relative low voltage level. Moreover, the transistors 22 and 30A are turned off. Due to the turned-on states of the transistors 13 and 30B, the transmission gate 31B passes the voltage VSS to the output node N11, thereby discharging the output node N11 toward the voltage VSS.
According to the embodiment of
According to the embodiment of
The N-type diffusion region 42 is also divided into four N-type diffusion portions 42A-42D by the polysilicon regions 46-48. The N-type diffusion portions 42A is disposed on the right side of the polysilicon region 46. The N-type diffusion portions 42B is disposed on the left side of the polysilicon region 46 and between the polysilicon regions 46 and 47. The N-type diffusion portions 42C is disposed on the right side of the polysilicon region 48 and between the polysilicon regions 47 and 48. The N-type diffusion portions 42D is disposed on the left side of the polysilicon region 48.
Two conductive segments 49 and 50 are formed above the P-type diffusion region 41 and the N-type diffusion region 42 and extend parallel to the direction D1. The conductive segments 49 and 50 are spaced apart from each other. Particularly, the P-type diffusion region 41 and the N-type diffusion region 42 are disposed between the conductive segments 49 and 50. The P-type diffusion region 41 is closed to the conductive segment 49, and the N-type diffusion region 42 is closed to the conductive segment 50.
A conductive segment 43 is formed under the P-type diffusion region 41 and the conductive segment 49 and extends parallel to the direction D2. A first end of the conductive segment 43 is electrically connected to the conductive segment 49 through a via V1, and a second end thereof is electrically connected to the P-type diffusion portion 41C. A conductive segment 44 is formed under the N-type diffusion region 42 and the conductive segment 50 and extends parallel to the direction D2. A first end of the conductive segment 44 is electrically connected to the conductive segment 50 through a via V2, and a second end thereof is electrically connected to the N-type diffusion portion 42C.
Referring to
A conductive segment 54 is formed under the P-type diffusion region 41 and the N-type diffusion region 42 and extends parallel to the direction D2. A conductive segment 51 is formed above the conductive segment 54 and the polysilicon regions 47 and 48 and extends parallel to the direction D1. The conductive segment 51 is disposed between the P-type diffusion region 41 and the N-type diffusion region 42. A first end of the conductive segment 54 is electrically connected to the P-type diffusion portion 41D, and a second end thereof is electrically connected to the N-type diffusion portion 42D. In other words, the conductive segment 54 electrically connects the P-type diffusion portion 41D to the N-type diffusion portion 42D. A first end of the conductive segment 51 is electrically connected to the conductive segment 54 through a via V7, and a second end thereof is electrically connected to the polysilicon region 47 through a via V8. In other words, the conductive segment 51 electrically connects the conductive segment 54 to the polysilicon region 47. Therefore, the P-type diffusion portion 41D, the N-type diffusion portion 42D, and the polysilicon region 47 are electrically connected together.
A conductive segment 52 is formed above the P-type diffusion region 41 and extends parallel to the direction D1. A conductive segment 56 is formed above the conductive segment 52 and extends parallel to the direction D2. A first end of the conductive segment 52 is electrically connected to the P-type diffusion portion 41A through a via V9. A first end of the conductive segment 56 is electrically connected to a second end of the conductive segment 52 through a via V10, and a second end thereof is electrically connected to the conductive segment 50 through a via V11.
A conductive segment 53 is formed above the N-type diffusion region 42 and extends parallel to the direction D1. A conductive segment 55 is formed above the conductive segment 53 and extends parallel to the direction D2. A first end of the conductive segment 53 is electrically connected to the N-type diffusion portion 42A through a via V12. A first end of the conductive segment 55 is electrically connected to a second end of the conductive segment 53 through a via V13, and a second end thereof is electrically connected to the conductive segment 49 through a via V14.
According to the above description, two regions are electrically connected through a via. For example, the conductive segment 43 is electrically connected to the conductive segment 49 through the via V1. However, in other embodiments, two regions may be electrically connected in another connection manner that depends upon the applied manufacturing process.
In the embodiment, the conductive segments 49-53 are formed on the same metal layer (Metal 0) that is close to the P-type diffusion region 41 and the N-type diffusion region 42. The conductive segments 55 and 56 are formed on the same metal layer (Metal 1) that is formed above the metal layer (Metal 0). The conductive segments 43-45 and 54 are formed on the same metal layer (Metal D) under the diffusion regions 41 and 42.
Referring to
The polysilicon region 47 and the P-type diffusion portions 41B-41C form the P-type transistor 30A of the inverter 11, and the polysilicon region 47 and the N-type diffusion portions 42B-42C form the N-type transistor 30B of the inverter 11. The polysilicon region 47, the P-type diffusion portion 41C, and the P-type diffusion portion 41B serve as the gate electrode, the first drain/source electrode, and the second drain/source electrode of the P-type transistor 30A. The polysilicon region 47, the N-type diffusion portion 42B, and the N-type diffusion portion 42C serve as the gate electrode, the first drain/source electrode, and the second drain/source electrode of the N-type transistor 30B.
The polysilicon region 46 and the P-type diffusion portions 41A-41B form the P-type transistor 23, and the polysilicon region 46 and the N-type diffusion portions 42A-42B form the N-type transistor 22. The polysilicon region 46, the P-type diffusion portion 41B, and the P-type diffusion portion 41A serve as the gate electrode, the first drain/source electrode, and the second drain/source electrode of the P-type transistor 23. The polysilicon region 46, the N-type diffusion portion 42A, and the N-type diffusion portion 42B serve as the gate electrode, the first drain/source electrode, and the second drain/source electrode of the N-type transistor 22.
The polysilicon region 46 is electrically connected to the input node N10. The input signal S10 from the input node N10 is applied to the polysilicon region 46. The conductive segments 49 is electrically connected to the voltage source terminal VS10 of the buffer circuit 3, and the conductive segments 50 is electrically connected to the voltage source terminal VS11 of the buffer circuit 3. The conductive segment 45 is electrically connected to the output node N11 of the buffer circuit 3.
While the invention has been described by way of example and in terms of the preferred embodiments, it is to be understood that the invention is not limited to the disclosed embodiments. To the contrary, it is intended to cover various modifications and similar arrangements (as would be apparent to those skilled in the art). Therefore, the scope of the appended claims should be accorded the broadest interpretation so as to encompass all such modifications and similar arrangements.
This application claims the benefit of U.S. Provisional Application No. 63/324,205, filed Mar. 28, 2022, the entirety of which is incorporated by reference herein.
| Number | Date | Country | |
|---|---|---|---|
| 63324205 | Mar 2022 | US |
