MOS FET With Current Sensing Terminal

The present invention relates to a power MOS FET and, more particularly, to such a MOS FET having a current sensing terminal.

A power MOS FET is usually constituted by a number of MOS FET cells which are connected in parallel to each other. It is frequently necessary to detect a current flowing through such a power MOS FET. Since such current may be increased, up to the order of 10 Amperes, it is not practical to measure such large current directly.

In order to solve this high current problem, a MOS FET having a current sensing terminal is disclosed in U.S. Pat. No. 4,553,084. In this patent, a current sensing element comprises a limited number of MOS FET cells connected in parallel with a power MOS FET. In such a MOS FET, having a current sensing terminal, the power MOS FET has a pair of source terminals. One of the source terminals is called a "Kelvin terminal" and the other to a "Miller terminal". A sense resistor having a value R is connected between the Miller terminal and the Kelvin terminal. A current I flowing through the power MOS FET can be defined by the following equation if the value R and a potential difference V across the sense resistor R are known:

where n is a ratio number between MOS FET cells constituting the power MOS FET and MOS FET cells constituting the current sensing element.

Assuming that n=3000, R=30 ohms and I=10A, the potential difference V of 100 mV arises at both ends of the sense resistor R.

In the conventional MOS FET having a current sense terminal, the sense resistor is provided externally of the power MOS FET. Therefore, the sense resistor tends to be considerably influenced by external noise, causing the current detection to be inaccurate.

For a large external noise, it may be possible to detect the noise and to exclude it from a measured result. However, a circuit construction for performing these procedures may become very complicated and the sensing efficiency may be also degraded.

An object of the present invention is to provide a MOS FET having a current sensing terminal which is capable of detecting current accurately.

Another object of the present invention is to provide a MOS FET having a current sensing terminal which is capable of detecting current efficiently.

A MOS FET having a current sensing terminal, according to the present invention, comprises a semiconductor pellet on which a number of MOS FET cells, preferably, vertical type MOS FET cells are formed integrally. A major number of MOS FET cells are connected in parallel with each other to constitute a power MOS FET the gate electrodes of the cells and the gate electrode of the current sensing element are connected to each other and to a common gate electrode wiring and gate pad. Source regions of all cells in the power MOS FET are connected to a source pad and a Kelvin pad. Source regions of the current sensing element are connected to a Miller pad. A resistive element for sensing current is disposed between the source electrode of the power MOS FET and the Miller pad. The resistive element includes, preferably, a film resistor made of material which is the same material as that of the gate electrode.

Thus, it is possible to integrate the sense resistor in the same semiconductor pellet. External noise can be shielded by a package in which the semiconductor pellet is mounted to remove the inaccuracy of current sensing due to external noise.

The above mentioned and other objects, features and advantages of this invention will becomes more apparent by reference to the following detailed description of the invention taken in conjunction with the accompanying drawings, wherein:

FIG. 1 is a circuit diagram of a MOS FET having a current sensing terminal, according to the present invention;

FIG. 2 is a schematic plan view of a semiconductor pellet for explaining a first embodiment of the present invention;

FIG. 3 is an enlarged cross sectin taken along a line A--A in FIG. 2;

FIG. 4 is an enlarged cross section taken along a line B--B in FIG. 2;

FIG. 5 is a schematic, enlarged plan view of a portion C in FIG. 2;

FIG. 6 is an enlarged cross section taken along a line Y--Y in FIG. 5;

FIGS. 7a to 7d are cross sections of a semiconductor pellet, showing manufacturing steps according to a first embodiment of the present manufacturing method;

FIG. 8 is a schematic plan view of a portion of a semiconductor pellet for explaining a second embodiment of the present invention; and

FIG. 9 is an enlarged cross section taken along a line Y--Y in FIG. 8.

A first embodiment of the present invention will be described with reference to FIGS. 1 to 6.

FIG. 1 is a circuit diagram of a MOS FET having a current sensing terminal, according to the present invention. The MOS FET comprises a MOS FET MN1 and a MOS FET MN2.

The MOS FET MN1 is a power MOS FET constituted by a plurality of MOS FET cells connected in parallel. A gate, a drain and a source of the MOS FET MN1 are connected to a gate terminal G, a drain terminal D and a source terminal S, respectively. The MOS FET MN2 is constituted by a single or a small number of MOS FET cells connected in parallel. A gate and a drain of the MOS FET MN2 are connected to the gate terminal G and the drain terminal D, respectively, and a source thereof is connected to a Miller terminal M. A sense resistor R is connected between a Kelvin terminal K connected to the source terminal S and the Miller terminal M. The sense resistor R is integrated on a semiconductor pellet on which the MOS FET MN1 and 2 are also integrated. In a conventional MOS FET having a current sensing terminal, such a sense resistor is not included as part of the integrated circuit, but is adaptively connected between a Kelvin terminal and a Miller terminal as an external component when required.

Referring to FIG. 2 which is a plan view of a semiconductor pellet 101 on which the MOS FETs MN1 and MN2 and the sense resistor R are integrated. A gate pad 102, gate electrode wirings 103 and 104, a Miller pad 106, a Kelvin pad 107 and a source pad 108 are also formed thereon. The semiconductor pellet 101 is mounted on a package (not shown). The gate pad 102, the Miller pad 106, the Kelvin pad 107 and the source pad 108 are wire-bonded to internal leads (not shown) of the package through which they are led to the gate terminal G, the Miller terminal M, the Kelvin terminal K and the source terminal S, respectively.

FIG. 3 is a cross section taken along a line A--A in FIG. 2. FIG. 4 is a cross sectin taken along a line B--B in FIG. 2. In FIGS. 3 and 4, the gate electrode wiring 103 and the gate pad 102 are made of aluminum and are in contact with an underlying polysilicon film 104b. The polysilicon film 104b is connected to a polysilicon film 104a constituting a mesh type gate electrode.

The source pad 108 is made of an aluminum wiring layer separated from the gate electrode wiring 103 and is connected to the source electrode 105.

The Kelvin pad 107 has a construction which is similar to the construction of the source pad 108 except that an area thereof is smaller than that of the source pad 108.

FIG. 5 is an enlarged plan view of a portion of FIG. 2 corresponding to the Miller pad 106. FIG. 6 is a cross section taken along a line Y--Y in FIG. 5. In FIGS. 5 and 6, the Miller pad 106 is an aluminum layer arranged in the same layer as the gate pad 102, the gate wiring 103, the Kelvin pad 107, the source electrode 105 and the source pad 108. The Miller pad 106 is separate from the source electrode 105 and is connected to a source 120 of the MOS FET MN2.

A film resistor 119 is formed between the source electrode 105 and the Miller pad 106. The film resistor 119 is a polysilicon film formed simultaneously with the formation of the gate electrode 104a and constitutes a sense resistor R.

In FIGS. 3, 4 and 6, an N type silicon substrate 111 may be epitaxially formed on a substrate whose resistivity is 2.0 ohm-cm. An N.sup.+ type layer 110 having resistivity of 0.008 to 0.018 ohm-cm is formed on the entire lower surface thereof. The drain electrode 109 is an aluminum film formed on the N.sup.+ layer 110. P.sup.+ type layers 112 are formed in an upper surface of the substrate 111 in portions which are below the gate pad 102, the gate wiring 103, a main portion of the Miller pad 106, the Kelvin pad 107 and the source pad 108, respectively. On each P.sup.+ type layer 112, a silicon oxide film 117 having thickness of 600 nm is formed. The polysilicon film 104b covers an area of the layer 112 corresponding to the gate pad 102 and the gate electrode wiring 103. A film resistor 119 of polysilicon is provided on the silicon oxide film 117 in an area of the Miller pad 106. These polysilicon films are covered by an insulating film 118 of such as phosphosilicate glass (PSG). The gate pad 102 and the gate electrode wiring 103 are connected through contact holes 121 formed in the insulating film 118 to the polysilicon film 104b. Similarly, the film resistor 119 is connected through a hole 122 to the source electrode 105 and through a hole 123 to the Miller pad 106. The Kelvin pad 107 and the source pad 108 are provided directly on the insulating film 118.

The mesh type gate electrode 104a is formed on a gate insulating film 116 of silicon oxide and of 50 nm thickness which is formed on the surface of the semiconductor substrate 111. Below each open portion of the mesh of the gate electrode 104a, a P type base region 114 is provided on the surface of the N type silicon substrate 111, as shown in FIGS. 4 and 6. In a center portion of the P type base region 114, a P.sup.+ type base region 113 is provided. The P type base region 114 extends up to a portion immediately below the gate electrode 104a. The N type source region 115 is provided in a portion of the P type base region 114 immediately below the gate electrode 104a, so as to surround the gate electrode. The source electrode 105 is connected through holes 124 of the insulating film 118 to an N type source region 115 and a P.sup.+ type base region 113 of the respective MOS FET cells.

As shown in FIGS. 3 and 4, an N type layer 115a is formed in an outer periphery of the semiconductor chip simultaneously with formation of the N type source region 115. The layer 115a is separated from the source electrode 105 and serves as a channel stopper connected to the aluminum wiring layer 105a in the same layer.

As to the size of the MOS FET, the thickness and width of the gate electrode 104a are 600 nm and 13 .mu.m, respectively. An area of the region defined by the P type base region 114 (square region shown in FIG. 5 by dotted lines) is 12 .mu.m.times.12 .mu.m. An area of the film resistor 119 is about 12 .mu.m.times.16 .mu.m, the resistance thereof being from 20 ohms to 30 ohms. A total number of MOS FET cells is about 3000. In this embodiment, the MOS FET MN2 is constituted with a single MOS FET cell.

The provision of the film resistor 119 between the source electrode 105 and the Miller pad 106 is the feature of the present invention over the conventional MOS FET having current sensing terminal.

As will be described later, there is no need for adding special steps for providing the film resistor 119. Further, the number of bonding pads is the same as the number of the conventional device or only four, i.e., the gate pad 102, the Miller pad 103, the Kelvin pad 107 and the source pad 108. The size of the semiconductor pellet is substantially the same as the size of the conventional ine.

A fabrication of the first embodiment shown in FIGS. 1 to 6 will be described with reference to FIG. 7.

As shown in FIG. 7(a), an N type silicon substrate 111 having an N.sup.+ type layer 110 formed on a lower surface thereof is prepared first. Then, a silicon oxide film 126 having a thickness of 750 nm is formed by thermal oxidation of the substrate. Then, the silicon oxide film is selectively removed to form openings. Boron is ion-implanted through the openings into the substrate and then diffused to form P type channel stopper 112, P.sup.+ type base region 113 of a MOS FET cell and field rings (not shown). The ion-implantation is performed with the energy of 70 KeV and the dose of 1.times.10.sup.14 /cm.sup.2, followed by a diffusion. Due to the diffusion after the ion-implantation, oxide films 127 are formed on surfaces of the P type regions 112 and 113.

Thereafter, as shown in FIG. 7(b), the silicon oxide films 126 and 127 in the cell region are removed and a gate insulating film 116 of silicon oxide having a thickness of 50 nm is formed. Alternatively, it may be possible to completely remove the silicon oxide films 126 and 127, in order to form a silicon oxide film having a thickness in the range from 100 nm to 1000 nm, selectively remove the silicon oxide film, and form a gate insulating film 116. On the surface of the channel stopper 112 outside the cell region, an oxide film is grown with a growth of the gate oxide film, resulting in a thick oxide film 117. A polysilicon film 128 having a thickness of 600 nm is formed over the entire surface.

Then, as shown in FIG. 7(c), the polysilicon film 128 is selectively removed to pattern the gate electrode 104a and the film resistor 119. Thereafter, the gate insulating film 116 is removed by using the thus patterned gate electrode 104a and the film resistor 119 as a mask. Then, boron is ion-implanted with the energy of 70 KeV and with a dose of 9.0.times.10.sup.13 /cm.sup.2. Then, a P type base region 114 is formed by Pa heat-treatment thereof. During this process, thin oxide film 129 is formed on the respective P type base region 114 and the P type base regions 113. Such a thin film is also formed on the gate electrode 104a and the film resistor 119.

Then, as shown in FIG. 7(d), an aluminum film of 1.2 .mu.m thickness is vapor-deposited. By selectively removing the aluminum film, an aluminum mask 130 is formed on the P.sup.+ type base region 113. Phosphorus is ion-implanted thereinto with an energy of 80 KeV and a dose of 5.times.10.sup.15 /cm.sup.2 and then heat-treated to form an N.sup.+ type source region 115.

Then, as shown in FIG. 6, the aluminum mask 130 is removed, an insulating film 118 of a material such as phosphosilicate glass is deposited. Then, holes 122, 123 and 124 are formed in the insulating film 118. Then, an aluminum film is vapor-deposited and patterned to form a gate pad 102, a gate electrode wiring 103, a source electrode 105, a mirror pad 106, a Kelvin pad 107 and a source pad 108.

As described, the film resistor 119 can be formed simultaneously with formation of the gate electrode 104a without the necessity of adding any special fabrication steps.

Although the present invention has been described as to an embodiment in which a single film resistor 119 is inserted between the source electrode 105 and the Miller pad 106, it may be possible to insert a desired number of film resistors in parallel and to trim them by laser irradiation. In the latter case, it is possible to improve the accuracy of the current sensing resistor.

A second embodiment of the present invention will be described with reference to FIGS. 8 and 9.

In FIGS. 8 and 9, a P type diffusion layer 131 is provided between a source electrode 105a and a Miller pad 106a. The P type diffusion layer 131 is formed simultanesouly with formation of a P type base region 114. The embodiment utilizes a diffusion resistor as the sense resistor. An absolute accuracy of the resistance thereof is about 15% which is advantageous when compared with 20% of the polysilicon resistor. Since other portions of the second embodiment are substantially the same as those of the first embodiment and are depicted by the same reference numerals as those used in FIGS. 5 and 6, the details thereof are omitted.

As described, the MOS FET having current sense terminal, according to the present invention, allows a load current to be sensed as a potential difference between the Miller terminal and the Kelvin terminal by providing the current sensing resistor within the semiconductor pellet. Therefore, the current sensing is not influenced by external noise. Further, since there is no need of providing any adaptive external resistor, the construction of the detector is simple and the detection efficiency can be improved. Further, there is no need for any additional step for providing the internal sensing resistor. There is no increase of the number of wire bonding and/or external terminals.

Although, in the two embodiments, the present invention is applied to a double-diffusion, vertical type MOS FET, the present invention can also be applicable to a V-MOS FET. Further, it should be noted that the positions of the source electrode and the drain electrode are not limited to the respective surfaces of the semiconductor pellet.

Although the present invention has been described with reference to specific embodiments, this description is not meant to be construed in a limiting sense. Various modifications of the disclosed embodiment, as well as other embodiments of the invention, will become apparent to persons skilled in the art upon reference to the description of the invention. It is therefore contemplated that the appended claims will cover any modifications or embodiments as fall within the true scope of the invention.