Inductor Formed at Least Partially in a Substrate

The following applications have been filed and may contain similar material as the present application, U.S. Ser. Nos. 08/363,145 and 08/355,750.

The invention relates to semiconductor devices and more particularly to an inductor formed in at least partially in a substrate of a semiconductor device.

Transformers amplify and attenuate fluctuating voltages. A transformer is formed with at least two inductors, a primary and a secondary, lying in close proximity (magnetically coupled) to each other. Each inductor consists of a core and a coil wound around the core. When a fluctuating voltage is applied to the coil of the primary inductor, magnetic lines of flux are created in the core of the primary inductor. The magnetic lines of flux permeate the core of the secondary inductor. The changing values of the magnetic lines of flux in the secondary inductor create a time varying current in the secondary coil and hence an induced emf (electromotive force). The induced emf in the secondary coil is directly proportional to the emf in the primary coil and the ratio of the number of turns of the secondary coil to the number of turns of the primary coil.

Typically, inductors have been large and heavy since they were made using a ferromagnetic core to enhance magnetic flux. Recently, inductors have been manufactured on a semiconductor wafer. One such inductor is described in U.S. Pat. No. 3,614,554.

The invention includes an inductor partially formed in the silicon in a silicon on insulator (SOI) structure or in another type of substrate of a semiconductor structure. The inductor of the invention is formed using a plurality of serially connected transistors having a common gate. The active area of the substrate used to form each transistor functions as the inductor coil, and the gate functions as the inductor core. Preferably, the gate is connected to the DC supply potential, Vcc.

In one application the invention includes a transformer comprised of two inductors of the invention. The transformer of the invention provides voltage control in order to increase or decrease an external supply potential, Vcc. The number of turns of the primary and secondary coils of the transformer, and thus the turns ratio of the transformer, is controllable by the number of transistors used in fabricating the primary and secondary inductors of the transformer.

In one currently envisioned embodiment a control transistor is serially connected to an external supply node and the primary coil of the transformer. The external supply node is connectable to Vcc. The control transistor is actuated intermittently by an output of a ring oscillator thereby creating a pulsating current in the primary coil. The pulsating current in the primary coil creates a changing magnetic flux in the primary core which permeates the secondary core thereby producing a corresponding pulsating current in the secondary coil. The pulsating current in the secondary coil is converted to obtain a DC potential.

FIG. 1A is a three dimensional view of a preferred embodiment of the invention.

FIG. 1B is a three dimensional view of a further preferred embodiment of the invention.

FIG. 1C is a simplified top planar view of the three dimensional views showed in FIGS. 1A and 1B.

FIG. 2A is a schematic of an embodiment of the invention.

FIG. 2B is a schematic of a further embodiment of the invention.

FIGS. 3-8 are cross sectional views of a method for forming the inductor of the invention.

FIG. 9 is a block schematic of a memory system of the invention.

FIG. 10 is a schematic of a transformer of the invention.

The invention includes an inductor. FIGS. 1A and 1B show a three dimensional cross section of two embodiments of the invention. A plurality of MOSFET transistors 5 are fabricated using a silicon-on-insulator (SOI) MOSFET process which uses the self isolation property of the SOI 10. Active areas 15 of the substrate are doped to create low resistance regions and/or source/drain regions. A channel is formed in each active area 15 during the actuation of its respective transistor 5. The active area 15 of each transistor 5 is surrounded by an oxide or other electrically insulative material 16 to electrically isolate it from active areas 15 of other transistors 5. The MOSFET transistors 5 are connected in series by meal connections 18 formed during a deposition and etch of at least one metal layer to overlie contacts 19 formed in contact with the source/drain regions of transistors 5. When the transistors 5 are activated, current flows in the active area 15 of the substrate in each transistor 5. The serially connected MOSFET transistors 5 form the coil of the inductor of the invention. Each transistor 5 comprises one turn of the coil. The MOSFET transistors 5 share a gate 20 which forms the core of the inductor. Preferably the gate 20 is connected to Vcc. This ensures that all of the serially connected transistors 5 are activated simultaneously, and that the coil of the inductor is continuous.

In FIG. 1A the active area 15 is an n+region overlying the oxide layer 25, which, in turn, overlies a p substrate 30. In this embodiment the entire film is opened rather than only the source/drain region. Alternately, the active area may be p- with the oxide layer 25 overlying an n type substrate. In addition other combinations used to make silicon on insulator substrates may be used. For example n+over n, p+over p, n+over p, and p+over n are a few of the options available to the designer.

In FIG. 1B the active area 15 is doped to form n+drain and source regions 35. In this embodiment the region underlying the gate 40 is p, and a p type substrate 30 underlies the oxide layer 25. Alternatively, the drain and source regions 35 may be p+ and the regions underlying the gate 40 and underlying the oxide 25 may be n.

FIG. 1C is a top planar view of the three dimensional views shown in FIGS. 1A and 1B, and the components have been numbered the same as in FIGS. 1A and 1B. A BPSG layer typically used to isolate interconnect 18 and contacts 19 from the gate is not shown in this view in order that the components of the inductor can be easily seen. A study of FIG. 1C reveals that the contacts 19, the interconnects 18 and the active regions 15 of the substrate form a conduction path around the gate 20 of the plurality of transistors, thereby allowing the gate 20 to function as a core of an inductor and allowing the conduction path formed in the plurality of active areas 15, contacts 19, and interconnects 18 to function as the coil of the inductor.

Referring now to FIG. 2A which is a schematic of one embodiment of the inductor 49 of the invention, the output signal of a ring oscillator 55 controls the actuation and deactuation of a control transistor 60 serially connected to the coil 50 and a supply node 65. The supply node 65 is connectable to the external supply potential, such as Vcc, or other DC potential. The core 70 of inductor 49 is connected to supply node 65. The core 70 is the common gate of serially connected transistors 72 which form the coil 50, and in this embodiment the gate, core 70, is continually biased when connected to a constant DC potential. The control transistor 60 actuates and deactuates in response to a periodic oscillator output potential. When the control transistor 60 is actuated serially connected transistors 72 conduct current. As the current changes in response to the actuation and deactuation of the control transistor 60 the magnetic lines of flux which permeate the core change. Thus, the change in magnetic flux is self-timed and dependent on the period and frequency of the periodic oscillator output potential.

In yet another embodiment, shown in FIG. 2B, the coil 50 of inductor 49 is connected between the supply node 65 and a reference node 71. A control transistor 60 is serially connected to supply node 65 and to the core 70 of inductor 49. The core 70 is actually the common gate of serially connected transistors 72 which form the coil 50. The periodic oscillator output potential of ring oscillator 55 controls the actuation and deactuation of the control is transistor 60. When control transistor 60 is actuated the supply potential is applied to the core 70 allowing transistors 72 to actuate and allowing current to flow in coil 50. When control transistor 60 is deactuated current flow ceases in the coil. As current varies in the coil 50 the magnetic lines of flux vary in the core 70.

It is also possible to connect a fluctuating potential directly to the coil 50 or the core 70. This can be done by connecting the output node of the ring oscillator 55 directly to the coil 50 or core 70.

In an alternate embodiment the core 70 and coil 50 are both connected to fluctuating potentials and timed to allow the actuation and deactuation of transistors 72.

The following is one currently envisioned embodiment for making the inductor of the invention. A blanket well implant is performed on an SOI wafer 100, see FIG. 3. The implant is followed by a well drive. Next pad oxide 105 is deposited, and a nitride layer 110 is deposited overlying the pad oxide 105. Active regions are defined by a mask 115, and the nitride layer 110, the pad oxide 105, and the silicon layer 120 are removed in unmasked regions, and oxide is deposited to create isolation regions 125 in openings formed during the removal of the nitride layer 110, the pad oxide 105, and the silicon layer 120, see FIG. 4.

Alternately, only the nitride layer 110 and the pad oxide 105 are removed. In this case the silicon remaining in the unmasked region is substantially consumed during a field oxide growth to create isolation regions 125. Following the formation of the isolation regions 125 the nitride layer 110 and the pad oxide 105 are removed in the active regions.

Sacrificial oxide is grown and the substrate is implanted to adjust the threshold voltage of the transistor being formed. This is referred to as a threshold voltage enhancement implant. The sacrificial oxide is then removed and gate oxide 130 is grown, see FIG. 5. Next a layer of polysilicon 135 is deposited and the gate is defined by a masking layer 140.

The polysilicon 135 and gate oxide 130 are removed in the unmasked regions, see FIG. 6. The polysilicon 135 remaining forms the core of the inductor of the invention and forms a transistor gate, in conjunction with the gate oxide 130, to the serially connected transistors which form the coil of the invention.

Following the formation of the gate or core a source drain implant is performed. Next BPSG 142 is formed according to conventional methods, and contacts 145 to the source/drain regions are formed in the BPSG using conventional methods, see FIG. 7. Next a metal layer 150 is deposited and masked with mask layer 155 to define interconnections between the drain and source for each two serially connected transistors of the plurality of transistors of the inductor of the invention. Please refer to FIG. 1C to enhance the interpretation of FIG. 7 three dimensionally.

In FIG. 8 the metal layer 150 has been removed in unmasked regions, and mask 155 has been removed. Thus a conduction path is formed with the contacts, interconnects and active areas of the transistors. The conduction path surrounds the core and forms the coil of the inductor of the invention. Please refer to FIG. 1C to enhance the interpretation of FIG. 8 three dimensionally.

FIG. 9 is a memory system 175 of one envisioned embodiment of the invention. The memory system 175 comprises a processor 180 and a monolithic memory device 185, such as, for example, a dynamic random access memory device. Two inductors 190 of the invention, as described above, are magnetically coupled and used in the monolithic memory device 185 as a transformer to attenuate or amplify the a DC potential. The processor 180 accesses the monolithic memory device 185.

FIG. 10 is a transformer 200 of the invention comprising a primary inductor 205 magnetically coupled by way of a common core 206 to a secondary inductor 210 of the invention. It can be seen from the schematic that the common core 206 is a gate common to transistors 211 and 212 which form the primary and secondary coils, respectively. By studying the schematic it can be seen that a mask having the shape of an annular ring can be used to fabricate the common core 206 of the transformer 200. Although one common gate has been shown in FIG. 10, it is also possible to fabricate the transformer wherein the cores of the primary and secondary inductors are physically separated. The periodic output of ring oscillator 215 controls the actuation and deactuation of control transistor 220 to generate a fluctuating current in transistors 211. Varying magnetic lines of flux are created in the primary inductor 205 in response to the fluctuating current generated in the primary inductor 205. The varying magnetic lines of flux permeate the secondary inductor 210 inducing a fluctuating current in the secondary inductor 210. A fluctuating potential is generated in the secondary inductor in response to the fluctuating current. A converter circuit 225 converts the fluctuating potential to a constant DC potential. This DC potential drives load 226, typically internal circuitry of a monolithic memory device. Thus, the transformer 200 attenuates, when the number of transistors 212 is less than the number of transistors 211, or amplifies, when the number of transistors 212 is greater than the number of transistors 211, the DC supply potential to internal circuitry of a monolithic memory device. In FIG. 10 the transistors 212 are electrically interposed between the converter circuit 225 and a reference potential at node 227. Alternately node 227 can be connected to a DC potential, such as Vcc. In either case the fluctuating potential generated in the secondary coil is superimposed on the value of the potential applied to node 227.

In addition, the designer or engineer skilled in the art can orient the coils in such a way to control the polarity of the output signal. Pages 398-423 and 694-700 of Basic Electronics, sixth edition, by Bernard Grob, McGraw-Hill Book Company are herein incorporated by reference.

In a further embodiment only one inductor of the invention is be used in combination with another type of inductor to form a transformer of the invention.

Although the invention has been described in preferred embodiments the invention should be limited only as claimed.