Analog circuit with improved layout for mismatch optimization

The present disclosure claims priority to U.S. Provisional Patent Application No. 61/857,943, filed on Jul. 24, 2013, which is incorporated herein by reference in its entirety.

Embodiments of the present disclosure relate to analog circuits, and in particular to optimizing mismatch between various components of an analog circuit.

Two substantially similar semiconductor components (e.g., two transistors) having substantially similar dimensions and configurations can have mismatch (e.g., a small difference in threshold voltages of the two transistors), for example, due to random manufacturing variability. Mismatch between various components in a semiconductor device can deteriorate the performance of the semiconductor device.

FIG. 1A schematically illustrates a conventional current mirror 100 comprising transistors Ta, Tb and Tc. In the current mirror 100 , a source of each of the transistors Ta, Tb and Tc is coupled to a supply voltage Vpdd. The gate of the transistors Ta, Tb and Tc are coupled to each other. Furthermore, the gate of the transistor Ta is coupled to a drain of the transistor Ta. A current Ipref output by the transistor Ta is mirrored as currents Ipouta and Ipoutb, respectively, through transistors Tb and Tc. For the currents Ipouta and Ipoutb to substantially mirror the current Iref, each of the transistors Tb and Tc have to be substantially matched to the transistor Ta.

FIG. 1B schematically illustrates a top view of the transistors Ta, Tb and Tc of FIG. 1A ; and FIG. 1C schematically illustrates a cross sectional view of the transistors Ta, Tb and Tc. As illustrated in FIGS. 1B and 1C , the transistor Ta has a gate region 14 a , a source region 16 a 1 and a drain region 16 a 2 formed on an N-well 12 a . Also illustrated in FIG. 1C is a pocket implant region 20 a (e.g., comprising N-type pocket implant dopants) formed near the source and drain regions of the transistor Ta. The transistors Tb and Tc have similar components as that of the transistor Ta.

The gate regions 14 a , 14 b and 14 c of the transistors Ta, Tb and Tc may have substantially similar dimensions (e.g., to reduce mismatch between the transistors). For example, in FIG. 1B , each of the gate regions 14 a , 14 b and 14 c of the transistors Ta, Tb and Tc have a length L and a width W.

Mismatch between the transistors Ta, Tb and Tc may be reduced, for example, by employing a relatively large width and/or large length of each of the transistors Ta, Tb and Tc. However, it may not always be feasible to employ a relatively large width and/or large length for the transistors Ta, Tb and Tc.

In a semiconductor device (e.g., especially in a semiconductor device in which the well or the substrate gets relatively lightly doped or almost undoped), a mismatch between various components is based on, for example, fluctuations of pocket implant dosage. As an example, a standard deviation of mismatch between threshold voltages of the transistors Ta, Tb and Tc is substantially proportional to an inverse of square root of an average number of dopants in the pocket implant regions of the transistors Ta, Tb and Tc. However, as illustrated in FIG. 1C , each of the transistors Ta, Tb and Tc has relatively few pocket implants. For example, dopants in the pocket implant region 20 a of transistor Ta is relatively low, compared to dimensions of the transistor Ta. Such low amount of dopants in the pocket implant regions may result in an increased mismatch between the transistors Ta, Tb and Tc, thereby deteriorating the performance of the current mirror 100 .

In various embodiments, the present disclosure provides a semiconductor device comprising: a substrate; a first transistor formed on the substrate; and a second transistor formed on the substrate, wherein a common region of the semiconductor device forms (i) a drain region of the first transistor, and (ii) a source region of the second transistor, and wherein a gate region of the first transistor is electrically coupled to a gate region of the second transistor. In an embodiment, a pocket implant region is formed on an edge of the common region; and the pocket implant region comprises pocket implant dopants. In an embodiment, a first region forms a source region of the first transistor; and a second region forms a drain region of the second transistor. In an embodiment, the common region is a first common region, and wherein the semiconductor device further comprises: a third transistor formed on the substrate, wherein a second common region of the semiconductor device forms (i) a drain region of the second transistor, and (ii) a source region of the third transistor, and wherein the gate region of the second transistor is electrically coupled to a gate region of the third transistor. In an embodiment, the gate region of the first transistor is electrically coupled to the gate region of the second transistor via a metal layer. In an embodiment, the first transistor and the second transistor are coupled in series. In an embodiment, the gate region of the first transistor has a first dimension and a second dimension; the gate region of the second transistor has a third dimension and a fourth dimension; the first dimension is substantially equal to the third dimension; the second dimension is substantially equal to the fourth dimension; and the first dimension and the third dimension are selected such that a sum of the first dimension and the third dimension is substantially equal to a pre-defined dimension. In an embodiment, the first transistor and the second transistor functionally operate as a single transistor. In an embodiment, each of the first transistor and the second transistor are metal oxide semiconductor field effect transistors (MOSFET).

In various embodiments, the present disclosure provides a method of forming a semiconductor device, the method comprising: forming a substrate; forming a first transistor on the substrate; and forming a second transistor on the substrate, wherein a common region of the semiconductor device forms (i) a drain region of the first transistor, and (ii) a source region of the second transistor, and wherein a gate region of the first transistor is electrically coupled to a gate region of the second transistor. In an embodiment, the method further comprises: forming a pocket implant region on an edge of the common region, wherein the pocket implant region comprises pocket implant dopants. In an embodiment, a first region forms a source region of the first transistor; and a second region forms a drain region of the second transistor. In an embodiment, the common region is a first common region, and wherein the method further comprises: forming a third transistor on the substrate, wherein a second common region of the semiconductor device forms (i) a drain region of the second transistor, and (ii) a source region of the third transistor, and wherein the gate region of the second transistor is electrically coupled to a gate region of the third transistor. In an embodiment, the gate region of the first transistor is electrically coupled to the gate region of the second transistor via a metal layer. In an embodiment, the first transistor and the second transistor are coupled in series. In an embodiment, the gate region of the first transistor has a first dimension and a second dimension; the gate region of the second transistor has a third dimension and a fourth dimension; the first dimension is substantially equal to the third dimension; the second dimension is substantially equal to the fourth dimension; and the method further comprises selecting the first dimension and the third dimension such that a sum of the first dimension and the third dimension is substantially equal to a pre-defined dimension. In an embodiment, the first transistor and the second transistor functionally operate as a single transistor. In an embodiment, each of the first transistor and the second transistor are metal oxide semiconductor field effect transistors (MOSFET).

Embodiments of the present disclosure will be readily understood by the following detailed description in conjunction with the accompanying drawings. To facilitate this description, like reference numerals designate like structural elements. Various embodiments are illustrated by way of example and not by way of limitation in the figures of the accompanying drawings.

FIG. 1A schematically illustrates a conventional current mirror comprising a plurality of transistors.

FIG. 1B schematically illustrates a top view of the plurality of transistors of FIG. 1A .

FIG. 1C schematically illustrates a cross sectional view of the plurality of transistors of FIG. 1A .

FIG. 2A schematically illustrates a current mirror comprising a plurality of transistors.

FIG. 2B schematically illustrates a top view of a subset of the plurality of transistors of FIG. 2A .

FIG. 2C schematically illustrates a cross sectional view of the subset of the plurality of transistors of FIG. 2A .

FIG. 3 is a flow diagram of a method for forming a semiconductor device.

FIG. 2A schematically illustrates a current mirror 200 comprising transistors Tp 1 , Tp 2 , Tp 3 , Tq 1 , Tq 2 , Tq 3 , Tr 1 , Tr 2 and Tr 3 . The current mirror 200 is similar to the current mirror 100 of FIG. 1A . However, in the current mirror 200 , the transistors Tp 1 , Tp 2 and Tp 3 replace the single transistor Ta of FIG. 1A . Similarly, in the current mirror 200 , the transistors Tq 1 , Tq 2 and Tq 3 replace the single transistor Tb of FIG. 1A ; and the transistors Tr 1 , Tr 2 and Tr 3 replace the single transistor Tc of FIG. 1A .

In an embodiment, the transistors Tp 1 , Tp 2 and Tp 3 are connected in series (henceforth also referred to as “a series of transistors”), such that a source of the transistor Tp 1 is coupled to a voltage source Vdd, a drain of the transistor Tp 1 is coupled to a source of the transistor Tp 2 , a drain of the transistor Tp 2 is coupled to a source of the transistor TP 3 , and a drain of the transistor Tp 3 is coupled to a current source Is that supplies a current Iref. A gate of the transistor Tp 1 is coupled to the gates of each of the transistors Tp 2 and Tp 3 . Furthermore, the drain of the transistor Tp 3 is coupled to the gates of each of the transistors Tp 1 , Tp 2 and Tp 3 .

The transistors Tq 1 , Tq 2 and Tq 3 are also connected in series, such that a source of the transistor Tq 1 is coupled to the voltage source Vdd, a drain of the transistor Tq 1 is coupled to a source of the transistor Tq 2 , a drain of the transistor Tq 2 is coupled to a source of the transistor Tq 3 , and a drain of the transistor Tq 3 outputs a current Iout 1 . A gate of the transistor Tq 1 is coupled to the gates of each of the transistors Tq 2 and Tq 3 . Furthermore, the gates of the transistors Tq 1 , Tq 2 and Tq 3 are coupled to the gates of the transistors Tp 1 , Tp 2 and Tp 3 , as illustrated in FIG. 2A .

The transistors Tr 1 , Tr 2 and Tr 3 are also connected in series in a manner that is similar to the connection of the transistors Tq 1 , Tq 2 and Tq 3 illustrated in FIG. 2A .

In an embodiment, the current Iref output by the transistor Tp 3 is mirrored as currents Iout 1 and Iout 2 , respectively, by the transistors Tq 3 and Tr 3 , as illustrated in FIG. 2A . For example, each of the currents Iout 1 and the current Iout 2 is substantially equal to the current Iref.

In an embodiment, the series of transistors Tp 1 , Tp 2 and Tp 3 is functionally equivalent to a single transistor Tp. For example, from a functional point of view, the transistors Tp 1 , Tp 2 and Tp 3 can be replaced by the single transistor Tp.

Similarly, from functional point of view, the transistors Tq 1 , Tq 2 and Tq 3 can be replaced by a corresponding single transistor Tq; and the transistors Tr 1 , Tr 2 and Tr 3 can be replaced by a corresponding single transistor Tr.

FIG. 2B schematically illustrates a top view of the transistors Tp 1 , Tp 2 and Tp 3 of the current mirror 200 of FIG. 2A ; and FIG. 2C schematically illustrates a cross sectional view of the transistors Tp 1 , Tp 2 and Tp 3 of the current mirror 200 of FIG. 2A . Although the transistors illustrated in FIGS. 2A-2C are P channel metal oxide semiconductor field effect (PMOS) transistors, the transistors can be of any other appropriate type (e.g., NMOS transistors). Although FIG. 2A illustrates the current Iref being mirrored as two currents Iout 1 and Iout 2 (i.e., the current mirror 200 including three series of transistors), in another embodiment, the current Iref may be mirrored as any different number of currents (i.e., the current mirror 200 may comprise any different number of mirrored paths, and any corresponding different number of series of transistors). Although FIGS. 2A-2C illustrate three corresponding transistors being coupled in a series of transistors (e.g., transistors Tp 1 , Tp 2 and Tp 3 being coupled in series), any different number of transistors (e.g., two, four, etc.) may also be coupled in a corresponding series of transistors.

As illustrated in FIGS. 2B and 2C , the transistors Tp 1 , Tp 2 and Tp 3 are formed on an N type substrate (e.g., N well 212 p ). The transistor Tp 1 has a corresponding gate region 214 p 1 . A region 216 pa , which is formed on the N-well 212 p and is doped with P type dopants, forms a source region of the transistor Tp 1 . A region 216 pb , which is formed on the N-well 212 p and is doped with P type dopants, forms a drain region of the transistor Tp 1 .

The transistor Tp 2 has a corresponding gate region 214 p 2 . Furthermore, the region 216 pb also forms a source region for the transistor Tp 2 . Thus, the region 216 pb is a common region for both the transistors Tp 1 and Tp 2 . A region 216 pc , which is formed on the N-well 212 p and is doped with P type dopants, forms a drain region of the transistor Tp 2 .

The transistor Tp 3 has a corresponding gate region 214 p 3 . Furthermore, the region 216 pc also forms a source region for the transistor Tp 3 . Thus, the region 216 pc is a common region for both the transistors Tp 2 and Tp 3 . A region 216 pd , which is formed on the N-well 212 p and is doped with P type dopants, forms a drain region of the transistor Tp 3 .

In an embodiment and although not illustrated in FIGS. 2B and 2C , the gate regions 214 p 1 , 214 p 2 and 241 p 3 of the transistors Tp 1 , Tp 2 and Tp 3 are coupled to each other. For example, the gate regions 214 p 1 , 214 p 2 and 241 p 3 are coupled via one or more metal layers, conductive traces, and/or the like.

Also illustrated in FIG. 2C are pocket implant regions 220 p formed near or on the edges of the regions 216 pa , 216 pb , 216 pc and 216 pd . In an embodiment, the pocket implant regions 220 p comprise N-type pocket implants or N type dopants.

In an embodiment, each of the gate regions 214 p 1 , 214 p 2 and 214 p 3 has a width of W 1 , and the gate regions 214 p 1 , 214 p 2 and 214 p 3 have lengths L 1 , L 2 and L 3 , respectively. In an embodiment, the lengths L 1 , L 2 and L 3 are substantially similar.

Although not illustrated in FIGS. 2B-2C , the series of transistors Tq 1 , Tq 2 , Tq 3 is structurally similar to the series of transistors Tp 1 , Tp 2 , Tp 3 ; and the series of transistors Tr 1 , Tr 2 , Tr 3 is structurally similar to the series of transistors Tp 1 , Tp 2 , Tp 3 . For example, the transistors Tq 1 , Tq 2 , Tq 3 have structures that are substantially similar to the structures of the transistors Tp 1 , Tp 2 , Tp 3 , respectively, as illustrated in FIGS. 2B and 2C . Accordingly, the structures of the series of transistors Tp 1 , Tp 2 , Tp 3 , and the series of transistors Tr 1 , Tr 2 , Tr 3 are not illustrated separately in the figures.

As previously discussed, the series of transistors Tp 1 , Tp 2 and Tp 3 is functionally equivalent to a single transistor Tp. For example, from a functional point of view, the transistors Tp 1 , Tp 2 and Tp 3 can be replaced by the single transistor Tp. In an example, a sum of the lengths L 1 , L 2 and L 3 of the transistors Tp 1 , Tp 2 and Tp 3 (i.e., an effective length of a gate of the functionally equivalent transistor Tp) is substantially equal to a predefined length. For example, if the transistors Tp 1 , Tp 2 and Tp 3 of the current mirror 200 of FIG. 2A are to replace the single transistor Ta of the current mirror 100 of FIG. 1A , then sum of the lengths L 1 , L 2 and L 3 of the transistors Tp 1 , Tp 2 and Tp 3 , for example, is substantially equal to the length L of the transistor Ta.

Replacing the transistor Ta in the current mirror 100 of FIGS. 1A-1C with the three transistors Tp 1 , Tp 2 and Tp 3 in the current mirror 200 of FIGS. 2A-2C has several advantages. For example, (i) the gate length of the transistor Ta and (ii) the effective gate length of the transistors Tp 1 , Tp 2 and Tp 3 are substantially similar. However, the transistor Ta has pocket implant regions 20 a formed near the regions 16 a 1 and 16 a 2 . In contrast, in the transistors Tp 1 , Tp 2 and Tp 3 , pocket implant regions 220 p are formed near regions 216 pa , 216 pb , 216 pc and 216 pd . Thus, a number of dopants in the pocket implant regions 220 p of the series of transistors Tp 1 , Tp 2 , Tp 3 is relatively higher than a number of dopants in the pocket implant regions 20 p of the transistor Ta. Thus, although the transistor Ta is functionally similar to the series of transistors Tp 1 , Tp 2 , Tp 3 , the series of transistors Tp 1 , Tp 2 , Tp 3 has a higher number of dopants in the pocket implant regions 220 p compared to that in the transistor Ta. Similarly, although the transistor Tb is functionally similar to the series of transistors Tq 1 , Tq 2 , Tq 3 , the series of transistors Tq 1 , Tq 2 , Tq 3 has a higher number of dopants in corresponding pocket implant regions compared to that in the transistor Tb. Similarly, although the transistor Tc is functionally similar to the series of transistors Tr 1 , Tr 2 , Tr 3 , the series of transistors Tr 1 , Tr 2 , Tr 3 has a higher number of dopants in corresponding pocket implant regions compared to that in the transistor Tc.

Increasing a number of dopants in pocket implant regions of two transistors results in a decrease in a standard deviation of a mismatch between threshold voltages of the two transistors. Thus, as the number of dopants in the pocket implant regions of each of the series of transistors Tp 1 , Tp 2 , Tp 3 , the series of transistors Tq 1 , Tq 2 , Tq 3 , and the series of transistors Tr 1 , Tr 2 , Tr 3 are relatively high (e.g., higher than the number of dopants in the pocket implant regions of each of the transistors Ta, Tb, Tc), a standard deviation of mismatch between threshold voltages of the equivalent transistors Tp, Tq and Tr is relatively low, thereby resulting in better matching of the currents Iref, Iout 1 and Iout 2 .

For example, assume that the transistor Ta has N number of dopants in the pocket implant regions 20 a . Accordingly, a standard deviation of mismatch between threshold voltages of the transistors Ta, Tb, Tc is substantially proportional to 1/sqrt(N), where sqrt(N) represents a square root of N. In the series of transistors Tp 1 , Tp 2 , Tp 3 , the number of dopants in the pocket implant regions 220 p is almost three times N, i.e., 3N (e.g., as the pocket implant regions 220 p is about thrice the size of the pocket implant regions 20 a ). Accordingly, a standard deviation of mismatch between the threshold voltages of the functionally equivalent transistors Tp, Tq, Tr is substantially proportional to 1/sqrt(3N), where sqrt(3N) represents a square root of 3N. Thus, the standard deviation of mismatch between threshold voltages associated with the series of transistors Tp 1 , Tp 2 , Tp 3 is much lower than that of the transistor Ta.

Thus, replacing a single transistor (e.g., transistor Ta of FIGS. 1A-1C ) with a functional equivalent series of transistors Tp 1 , Tp 2 and Tp 3 results in an improvement in the mismatch, without substantially increasing effective dimensions of the gate region (i.e., as discussed, dimensions of the gate region 14 a of the transistor Ta is substantially equal to effective dimensions of the gate regions 214 p 1 , 214 p 2 , 214 p 3 of the transistors Tp 1 , Tp 2 , Tp 3 ).

In an embodiment, having a series of transistors Tp 1 , Tp 2 and Tp 3 in the current mirror 200 of FIGS. 2A-2C (e.g., instead of a single transistor) also ensures, for example, that a line edge roughness (LER) effect along a gate length in the series of transistors direction is averaged out due to multiple stacked transistors, thereby making a critical dimension of the gate region of the transistors a relatively smaller factor in device mismatch.

As previously discussed, in an embodiment, the lengths L 1 , L 2 and L 3 of the gate regions 214 p 1 , 214 p 2 and 214 p 3 are substantially equal (e.g., say equal to a length Lp). In an embodiment, the length Lp is equal or near equal to a minimum length of a gate region, as allowed for a given technology used to fabricate the transistors of the current mirror 200 . This allows usage of a maximum number of transistors in a series of transistors.

For example, assume that the gate length L of the transistor Ta is 100 units. The transistor Ta can be replaced by a series of transistors, where the series of transistors comprises two, three, four, or more number of transistors (e.g., the series of transistors in the example of FIGS. 2A-2C comprises three transistors tp 1 , tp 2 , tp 3 ). A first option would be to replace the transistor Ta with two transistors, each having a gate length of 50 units; a second option would be to replace the transistor Ta with three transistors, each having a gate length of 33.33 units; a third option would be to replace the transistor Ta with four transistors, each having a gate length of 25 units; and so on. Also assume that the technology used for fabricating the transistors allows a minimum gate length of 30 units. As the minimum gate length has to be 30 units, the third option of replacing the transistor Ta with four transistors cannot be implemented. In such a case, the second option is to be chosen, in which the transistor Ta is replaced with three transistors, each having a gate length of 33.33 units. Choosing the second option, instead of the first option, allows a maximum number of transistors (i.e., three) used for replacing the transistor Ta (e.g., instead of using two transistors to replace the transistor Ta), thereby increasing an amount of dopant in the pocket implant regions, and thus decreasing possibilities of mismatch.

In an embodiment, a distance between gates of two adjacent transistors in a series of transistors (e.g., distance between gate regions 214 p 1 and 214 p 2 in FIGS. 2B and 2C ) is made as small as possible, as allowed under the technology used to fabricate the transistors. Having smaller distance between gates of two adjacent transistors in a series of transistors, for example, minimizes parasitic resistances and improves layout of the transistors (e.g., by reducing an area occupied by the series of transistors).

Although FIG. 2A illustrates using the various series of transistors in a current mirror, the various series of transistors (e.g., the series of transistors Tp 1 , Tp 2 , Tp 3 , the series of transistors Tq 1 , Tq 2 , Tq 3 , and/or series of transistors Tr 1 , Tr 2 , Tr 3 ) can be used in any other appropriate circuit as well.

FIG. 3 is a flow diagram of a method 300 for forming a semiconductor device (e.g., the series of transistors Tp 1 , Tp 2 , Tp 3 ). At 304 , a substrate (e.g, the N well 212 p of FIGS. 2B and 2C ) is formed. At 308 , a first transistor (e.g., the transistor Tp 1 of FIGS. 2A-2C ) is formed on the substrate. At 312 , a second transistor (e.g., the transistor Tp 2 of FIGS. 2A-2C ) is formed on the substrate. In an embodiment, a common region (e.g., the region 216 pb of FIG. 2C ) of the semiconductor device forms (i) a drain region of the first transistor, and (ii) a source region of the second transistor. In an embodiment, a gate region (e.g., the gate region 214 p 1 ) of the first transistor is electrically coupled to a gate region (e.g., the gate region 214 p 2 ) of the second transistor.

The description may use the phrases “in an embodiment,” or “in embodiments,” which may each refer to one or more of the same or different embodiments. The phrase “in some embodiments” is used repeatedly. The phrase generally does not refer to the same embodiments; however, it may. The terms “comprising,” “having,” and “including” are synonymous, unless the context dictates otherwise. The phrase “A and/or B” means (A), (B), or (A and B). The phrase “A/B” means (A), (B), or (A and B), similar to the phrase “A and/or B.” The phrase “at least one of A, B and C” means (A), (B), (C), (A and B), (A and C), (B and C) or (A, B and C). The phrase “(A) B” means (B) or (A and B), that is, A is optional.

Although certain embodiments have been illustrated and described herein, a wide variety of alternate and/or equivalent embodiments or implementations calculated to achieve the same purposes may be substituted for the embodiments illustrated and described without departing from the scope of the present invention. This application is intended to cover any adaptations or variations of the embodiments discussed herein. Therefore, it is manifestly intended that embodiments in accordance with the present invention be limited only by the claims and the equivalents thereof.