SUBTHRESHOLD CMOS INTEGRATOR

Abstract

An integrator circuit having a relatively large RC time constant includes a resistive element implemented with a field effect transistor operated in a sub-threshold mode. The size of the field effect transistor is selected, in addition to the sub-threshold gate voltage, to achieve a desired resistance value in a small area and without using bipolar devices. A differential integrator circuit includes two field effect transistors operated in a sub-threshold mode, with a capacitor connected between the output terminals of the two field effect transistors. A bulk drive circuit can be optionally used to reduce high frequency in the bulk.

Description

FIELD OF THE INVENTION

[0001] The present invention relates to electronic circuits and, more particularly, to integrator circuits.

[0002] BACKGROUND INFORMATION

[0003] FIG. 1 illustrates a conventional active RC integrator circuit 10 having an operational amplifier 11, a capacitor 12 and a resistor 13 with resistance R. Resistor 13 has one electrode connected to the inverting input terminal of operational amplifier 11 and the other electrode connected to receive an input signal VIN. Capacitor 12 is connected between the output lead and inverting input terminal of operational amplifier 11. The non-inverting input terminal of operational amplifier 11 is connected to ground. In operation, input signal VIN is converted to a current by resistor 13, which is then integrated by capacitor 12 to provide an output signal VOUT. Such integration circuits are commonly used in a wide variety of applications and may be tuned by manipulating the RC time constant of the feedback loop to operate over a range of frequencies.

[0004] In low frequency applications, the RC time constant of the feedback loop must be relatively large in order for the integrator to function properly. For example, in an audio application, the RC time constant is typically on the order of eight milliseconds. Further, the capacitor should be small to minimize the cost of fabricating a monolithic integrated circuit integrator circuit, typically having a capacitance no greater than twenty picofarads. Consequently, the resistor must have a resistance on the order of four hundred mega ohms. Such a large resistor can be fabricated, but conventional resistive elements in integrator applications generally require a relatively large amount of semiconductor real estate, which tends to increase fabrication costs.

SUMMARY

[0005] In accordance with aspects of the present invention, an integrator circuit having a relatively large RC time constant is provided. In one aspect of the present invention, the integrator circuit includes a resistive element implemented with a field effect transistor operated in a sub-threshold mode. By controlling the size of the field effect transistor in addition to the sub-threshold gate voltage, a desired resistance value can be achieved in a small area and without using bipolar devices, thereby decreasing fabrication costs and complexity.

[0006] In a further aspect of the present invention, the integrator circuit is fully differential with the integrating capacitor being implemented as a floating capacitor. In accordance with this aspect of the invention, the integrator circuit is only responsive to differential signals and allows integration to take place on both sides of the floating capacitor.

[0007] In another aspect of the present invention, the integrator circuit also includes a bulk drive circuit that uses the differential input signal to bias the bulk keep the bulk quiet. More particularly, the bulk drive circuit uses gate capacitance of matching bulk drive transistors to cancel high frequency noise.

BRIEF DESCRIPTION OF THE DRAWINGS

[0008] FIG. 1 illustrates a block diagram of a conventional integrator circuit.

[0009] FIG. 2 illustrates a block diagram of an integrator circuit having a resistor implemented with a field effect transistor biased in the sub-threshold mode, in accordance with one embodiment of the present invention.

[0010] FIG. 3 illustrates a chart of current as a function of drain-source voltage of an exemplary PMOS device, showing a ten fold increase in saturation current per 120 mV change in gate voltage in sub-threshold mode.

[0011] FIG. 4 illustrates a chart of current as a function of drain-source voltage of exemplary NMOS devices having different channel lengths but the same gate voltage.

[0012] FIG. 5 illustrates a block diagram of a differential integrator circuit having a resistor implemented with a field effect transistor biased in the sub-threshold mode, in accordance with one embodiment of the present invention.

[0013] FIG. 6 illustrates a block diagram of a differential integrator circuit as in FIG. 5 with the addition of a bulk drive circuit, in accordance with one embodiment of the present invention.

[0014] FIG. 7 illustrates a circuit diagram of one implementation of the integrator circuit of FIG. 6, according to one embodiment of the present invention.

DETAILED DESCRIPTION

[0015] FIG. 2 illustrates an integrator circuit 20 according to one embodiment of the present invention. Integrator circuit 20 includes a transconductance amplifier resistor implemented with a field effect transistor biased in the sub-threshold mode, in accordance with one embodiment of the present invention. This embodiment is similar to integrator circuit 10 (FIG. 1) except that resistor 13 is implemented with a field effect transistor 23 that is biased by a bias circuit 25 to operate in the sub-threshold mode. By operating in the subthreshold mode, field effect transistor 23 operates as a resistor with high resistance. Subthreshold mode resistance performance is described below in conjunction with FIGS. 3 and 4. Transistor 23 and bias circuit 25 require relatively little silicon real estate and dissipate little power in achieving this high resistance. In addition, transistor 23 and bias circuit 25 can be implemented without using bipolar devices, which can advantageously reduce the complexity and cost of fabricating integrator circuit 20.

[0016] FIG. 3 illustrates a chart of current as a function of drain-source voltage of an exemplary PMOS process. In this example, a minimum geometry PMOS device (e.g., having a channel width and length of &mgr;m each) is operated in the subthreshold mode with gate voltages that are incremented by 120 mV. As shown in FIG. 3, the gate voltages are 680 mV, 800 mV and 920 mV. Each 120 mV increase results in a ten-fold increase in drain saturation current. At low current densities (typically below a nanoAmp for minimum geometry devices), this exemplary PMOS device exhibits a resistance of approximately 52 mV/Idsat when operated in the subthreshold mode, where Idsat represents the level of the drain saturation current.

[0017] FIG. 4 illustrates a chart of current as a function of drain-source voltage of an exemplary NMOS process. In this example, the NMOS devices have a gate voltage of 0.7 volts and a width of 3 &mgr;m, with the length ranging from 1 &mgr;m to 100 &mgr;m. FIG. 4 shows that altering the width and length of the NMOS device scales the drain saturation current for reasonably sized transistors. In view of the present disclosure, those skilled in the art of integrator circuits can use charts similar to those of FIGS. 3 and 4 (i.e. determined from testing devices of the process to be used in fabricating the integrator circuit) to implement a resistor with a high resistance (i.e., over 100 M&OHgr;). In particular, by selecting the size and the subthreshold biasing of a field effect transistor, the designer can achieve a desired resistance.

[0018] FIG. 5 illustrates a block diagram of a differential integrator circuit 50, in accordance with one embodiment of the present invention. Differential integrator circuit 50 includes a differential amplifier 51, resistors 521 and 522, and capacitor 12. Resistor 521 is implemented with a field effect transistor 531 biased in the sub-threshold mode using a bias circuit 551. Similarly, resistor 522 is implemented with a field effect transistor 532 biased in the sub-threshold mode using a bias circuit 552. In one embodiment, capacitor 12 is implemented with two MOS capacitors connected in parallel. This allows capacitor 12 to have equal leakage on its terminals so that the differential output signal is not offset over time. Field effect transistors 531 and 532 and bias circuits 551 and 552 are fabricated, ideally, to be identical, with bias circuits 551 and 552 sinking identical bias currents.

[0019] The elements of differential integrator circuit 50 are interconnected as follows. Differential amplifier 51 has a first input lead connected to receive a signal VIN1 and a second input lead connected to receive a signal VIN2. Signals VIN1 and VIN2 form a differential input signal. Differential amplifier 51 outputs a differential output signal through lines 541 and 542. Line 541 is connected to one channel terminal of field effect transistor 531 and bias circuit 551, with bias circuit 551 also being connected to the gate of field effect transistor 531. The other channel terminal of field effect transistor 531 is connected to an electrode (i.e., node 561) of capacitor 12. Similarly, line 542 is connected to one channel terminal of field effect transistor 532 and bias circuit 552, with bias circuit 552 also being connected to the gate of field effect transistor 532. The other channel terminal of field effect transistor 532 is connected to the other electrode (i.e., node 562) of capacitor 12. The voltage across capacitor 12 serves as the differential output signal of integrator circuit 50.

[0020] In operation, differential integrator 50 is fully differential, being non-responsive to common mode voltages and currents. Further, the symmetry of the circuit cause any change in voltage or current on one side of capacitor 12 to result in an equal but opposite change in voltage or current on the other side of capacitor 12. Consequently, the average or common mode voltage across capacitor 12 remains constant; i.e., capacitor 12 floats.

[0021] Differential integrator 50 performs the integration function as follows. When the levels of signals VIN1 and VIN2 are equal, differential amplifier 51 outputs equal voltages to lines 541 and 542. The only DC current paths to the electrodes of capacitor 12 are through field effect transistors 531 and 532 on to lines 541 and 542. Therefore, no voltage appears across capacitor 12.

[0022] However, when the levels of signals VIN1 and VIN2 are not equal, differential amplifier 51 causes unequal voltages and currents to be present on lines 541 and 542. For example, when the differential input signal causes the voltage level on line 541 to drop (and conversely, the voltage level on line 542 to increase), field effect transistor 531 conducts less current (and conversely, field effect transistor 552 conducts more current). Consequently, more current flows through node 562 than through node 561, thereby causing current to flow through capacitor 12 from node 562 to node 561. The resultant voltage appearing across capacitor 12 forces the voltage at node 562 to increase and that at node 561 to decrease. Further, the symmetrical design causes the increase in voltage at node 562 to be matched by an equal decrease in voltage at node 561. The voltage across capacitor 12 is dependent on the integral of the differential input voltage (i.e., VIN1-VIN2). As previously described, field effect transistors 531 and 532 are sized and biased to operate in the sub-threshold mode to implement a relatively large resistance. For example, with a length ranging from 100 &mgr;m to 2000 &mgr;m and a width ranging from 3 &mgr;m to 10 &mgr;m and biased from a transistor that conducts a saturation current of 1 &mgr;A or lower, resistances well over 100 M&OHgr; can be achieved. This relatively high resistance can be achieved in a relatively small area of ranging from 100 &mgr;m×100 &mgr;m to about 500 &mgr;m×500 &mgr;m and allows capacitor 12 to be relatively small with a capacitance ranging from 5 pF to 20 pF. Further, biasing the field effect transistors requires relatively little power, on the order of 10 &mgr;W. This small size and power dissipation can be a tremendous advantage in low voltage battery-powered applications. Still further, field effect transistors 531 and 532 and bias circuits 551 and 552 can be implemented in a pure CMOS, thereby reducing complexity and cost relative to a design that requires bipolar devices.

[0023] FIG. 6 illustrates a block diagram of a differential integrator circuit 60, according to another embodiment of the present invention. Differential integrator circuit 60 is substantially similar to differential integrator circuit 50 (FIG. 5), but with the addition of a bulk drive circuit 62. Bulk drive circuit 62 is connected to lines 541 and 542, which allows bulk drive circuit 62 to receive differential mode noise. Bulk drive circuit 62 is also connected to the bulks of field effect transistors 531 and 532. Bulk drive circuit 62 includes capacitance (not shown) that operates to cancel high frequency noise. Otherwise, differential integrator circuit 60 operates as described above for differential integrator circuit 50 (FIG. 5). One embodiment of bulk drive circuit 62 is described below in conjunction with FIG. 7.

[0024] FIG. 7 illustrates a circuit diagram of one implementation of the integrator circuit 60 (FIG. 6), according to one embodiment of the present invention. In this embodiment, differential amplifier 51 is implemented with resistors 711 and 712, source coupled N-channel transistors 721 and 722 and a current source 73. Resistors 711 and 712 connect the VCC bus to the drains of N-channel transistors 721 and 722, respectively. Current source 73 is connected to the sources of N-channel transistors 721 and 722, which in turn have their gates connected to receive signals VIN1 and VIN2, respectively. N-channel transistors 721 and 722 also have their drains connected to lines 541 and 542, respectively.

[0025] Bulk drive circuit 62 is implemented using field-effect transistors 741 and 742, which are sized to match field effect transistors 531 and 532. In this embodiment, field effect transistors 741, 742, 531 and 532 are P-channel MOSFETs with a width of 3 &mgr;m and a length of 400 &mgr;m. As previously described, in other embodiments the sizes of these transistors may range from a width of 3 &mgr;m to 10 &mgr;m and a length of 100 &mgr;m to 2000 &mgr;m to achieve resistances from 100 M&OHgr; to well over 3000 M&OHgr; when operated in the sub-threshold mode. Field effect transistor 741 has one channel terminal connected to (or implemented by) the bulk and its other channel terminal connected to line 541. Similarly, field effect transistor 742 has one channel terminal connected to (or implemented by) the bulk and its other channel terminal connected to line 542. The gates of field effect transistors 741 and 742 are connected to the gates of field effect transistors 532 and 531, respectively. Field effect transistors 741 and 742 implement a differential bulk drive circuit having capacitance (i.e., the gate capacitance of the transistors) cancels high frequency noise injected into the bulk providing an equal and opposite signal that is also coupled into the bulk.

[0026] Bias circuits 551 and 552 are implemented using MOS diode 751 and current source 771, and MOS diode 752 and current source 772, respectively. MOS diode 751 has its anode connected to line 541 and has its cathode connected to current source 771 and the gate of field effect transistor 531. Similarly, MOS diode 752 has its anode connected to line 542 and has its cathode connected to current source 772 and the gate of field effect transistor 532. The sizes of the MOS diodes and the current sources are selected to achieve a gate-to-source voltage that biases field effect transistors 531, 532, 741 and 742 in the sub-threshold mode to achieve a desired resistance. In this embodiment, MOS diodes 751 and 752 are implemented with matching P-channel field effect transistors having a width of 400 &mgr;m and a length of 3 &mgr;m. Also, current sources 771 and 772 are ideally matching current sources, which in this embodiment are each configured to sink a 1 &mgr;A current. In other embodiments, the sizes of field effect transistors 751 and 752 and current sources 771 and 772 can be different as needed to properly bias field effect transistors 531, 532, 741 and 742 in the sub-threshold mode to achieve different desired resistances.

[0027] Differential integrator 70 performs the integration function as follows. When the levels of signals VIN1 and VIN2 are equal, field effect transistors 721 and 722 conduct equal currents that result in equal voltages at lines 541 and 542. Since lines 541 and 542 provide (through field effect transistors 531 and 532) the only current path to the electrodes of capacitor 12, the voltage across capacitor 12 is therefore equal.

[0028] However, when the levels of signals VIN1 and VIN2 are not equal, field effect transistors 721 and 722 of differential amplifier 51 conduct unequal currents. This unequal current flow causes unequal voltages at lines 541 and 542. Since these two lines provide (through field effect transistors 531 and 532) the only current path to the electrodes of capacitor 12, the voltage across capacitor 12 is therefore as unequal by the same average amount. Because of the high resistances of field effect transistors 531 and 532, the variation in voltages at lines 541 and 542 will get integrated over time by capacitor 12. The voltage across capacitor 12 is dependent on the integral of the differential input voltage (i.e., VIN1-VIN2).

[0029] In a further refinement, the sizes of field effect transistors 531, 532, 741, and 742 and current sources 771 and 772 may be programmable. For example, additional transistors may be programmably connected in parallel to field effect transistor 531 to increase its size. In this way, the resistance provided by field effect transistor 531 may be trimmed as desired to account for process variations. This programming can be performed using mask programmable techniques, laser trimming, fuse or antifuse programming, or non-volatile memory.

[0030] The above specification, examples and data provide a complete description of the manufacture and use of the composition of the invention. In light of the present disclosure, many embodiments of the invention can be made without departing from the spirit and scope of the invention by those skilled in the art of integrator circuits. For example, although a PMOS resistor implementation is described, those skilled in the art can fabricate a NMOS resistor implementation without undue experimentation. Accordingly, the invention is not to be limited to those embodiments disclosed, but rather, the invention resides in the claims hereinafter appended.

Claims

1. An integrator circuit comprising:

a differential amplifier;

a first field effect transistor coupled to a first output terminal of the differential amplifier;

a second field effect transistor coupled to a second output terminal of the differential amplifier

a first bias circuit coupled to the first field effect transistor, wherein the first bias circuit is configured to bias the first field effect transistor in a sub-threshold mode;

a second bias circuit coupled to the second field effect transistor, wherein the second bias circuit is configured to bias the second field effect transistor in a sub-threshold mode; and

a capacitor having a first electrode coupled to the first field effect transistor and having a second electrode coupled to the second field effect transistor.

2. The integrator circuit of claim 1 further comprising a bulk drive circuit coupled to a bulk, the first and second field effect transistors being formed in the bulk, wherein the bulk drive circuit is configured to cancel noise injected into the bulk.

3. The integrator circuit of claim 2 wherein the bulk drive circuit comprises third and fourth field effect transistors, the third field effect transistor being coupled to the first output terminal of the differential amplifier and to the gate of the second field effect transistor, the fourth field effect transistor being coupled to the second output terminal of the differential amplifier and to the gate of the first field effect transistor.

4. The integrator circuit of claim 1 wherein the first bias circuit comprises a constant current source.

5. The integrator circuit of claim 4 wherein the constant current source is programmable.

6. The integrator circuit of claim 4 wherein the first bias circuit further comprises a transistor coupled to the first output terminal and to the constant current source.

7. The integrator circuit of claim 1 wherein the sizes of the first and second field effect transistors are programmable.

8. The integrator circuit of claim 1 wherein the capacitor comprises matching MOS capacitors connected in parallel such that leakage associated with the first electrode of the capacitor matches leakage associated with the second electrode of the capacitor.

9. The integrator circuit of claim 1 wherein the capacitor has a capacitance ranging from 5 pF to 20 pF.

10. The integrator circuit of claim 1 wherein the first field effect transistor has a width ranging from 3 &mgr;m to 10 &mgr;m and a length ranging from 100 &mgr;m to 2000 &mgr;m.

11. An integrator circuit comprising:

a differential amplifier;

a first field effect transistor coupled to a first output terminal of the differential amplifier;

a second field effect transistor coupled to a second output terminal of the differential amplifier

first bias means, coupled to the first field effect transistor, for biasing the first field effect transistor in a sub-threshold mode;

second bias means, coupled to the second field effect transistor, for biasing the second field effect transistor in a sub-threshold mode; and

a capacitor having a first electrode coupled to the first field effect transistor and having a second electrode coupled to the second field effect transistor.

12. The integrator circuit of claim 11 further comprising a bulk drive means, coupled to a bulk, for canceling noise injected into the bulk, the first and second field effect transistors being formed in the bulk.

13. The integrator circuit of claim 12 wherein the bulk drive means comprises third and fourth field effect transistors, the third field effect transistor being coupled to the first output terminal of the differential amplifier and to the gate of the second field effect transistor, the fourth field effect transistor being coupled to the second output terminal of the differential amplifier and to the gate of the first field effect transistor.

14. The integrator circuit of claim 11 wherein the first bias means comprises a constant current source.

15. The integrator circuit of claim 14 wherein the constant current source is programmable.

16. The integrator circuit of claim 14 wherein the first bias means further comprises a transistor coupled to the first output terminal and to the constant current source.

17. The integrator circuit of claim 11 wherein the sizes of the first and second field effect transistors are programmable.

18. The integrator circuit of claim 11 wherein the capacitor comprises matching MOS capacitors connected in parallel such that leakage associated with the first electrode of the capacitor matches leakage associated with the second electrode of the capacitor.

19. The integrator circuit of claim 11 wherein the capacitor has a capacitance ranging from 5 pF to 20 pF.

20. The integrator circuit of claim 11 wherein the first field effect transistor has a width ranging from 3 &mgr;m to 10 &mgr;m and a length ranging from 100 &mgr;m to 2000 &mgr;m.