WO2007075617A2

WO2007075617A2 - Current sensor

Info

Publication number: WO2007075617A2
Application number: PCT/US2006/048325
Authority: WO
Inventors: Timothy J. Dupuis; John B. Pavelka; Vivek Sarda
Original assignee: Silicon Laboratories Inc.
Priority date: 2005-12-19
Filing date: 2006-12-19
Publication date: 2007-07-05
Also published as: WO2007075617A3

Abstract

A current sensor includes one or more inductors that generate a sensed output current responsive to a current in a conductor when the one or more inductors are inductively coupled to the conductor. The current sensor includes an integrated circuit die including an integrator circuit coupled to the one or more inductors for generating a sensed voltage responsive to the sensed current, the sensed voltage indicative of the first current in the conductor. The inductors may be formed in the integrated circuit and the conductor may incorporated into the package holding the die.

Description

CURRENT SENSOR

TECHNICAL FIELD

[1001] The present invention relates to current sensors and more particularly to an improved current sensor having low loss., small size, low cost, and high accuracy.

BACKGROUND ART

[1002] Within various circuit implementations, such as power supplies, there is often a need to detect a current provided at a particular point within a circuit. For example, a detected current may be used as feedback for controlling other parts of a circuit. Various techniques are presently used to sense currents within electronic circuits, but each of these techniques has shortcomings. One approach, illustrated in Fig. 1, utilizes a resistor 102 connected across the inputs of an operational amplifier 104 to provide a voltage V_SENSE that may be used to determine a current 106. A low value resistor, in the range of 10 mOhms, may be implemented. However, a drawback of this approach is the high loss provided by the circuit. The high loss may be mitigated by reducing the value of the resistor 102, however, this also reduces the signal V_SENSE that maybe detected. While this type of circuit may be used to sense current in direct current (DC) applications, the resistor 102 has generally not been capable of being reaφly integrated on an integrated circuit.

[1003] Referring now to Fig. 2, a further prior art system, utilizing a Hall effect device 202, connected across the inputs of an operational amplifier 204, is illustrated. The Hall effect device 202 generates a voltage across the inputs of the operational amplifier 204, responsive to a current 206, to provide an output signal V_SENSE- While this approach has a relatively low loss and may be used to detect direct current (DC), the use of the Hall effect device 202 generally provides a circuit having a higher cost. Furthermore, accuracy and noise issues are generally greater in current sensors that implement a Hall device, as the Hall voltage is a relatively small value.

[1004] With reference to Fig. 3, a current sensor that uses a magneto resistive sensor is illustrated. The magneto resistive sensor consists of a magneto resistive element 302 connected across the inputs of operational amplifier 304 to detect a current 306. The magneto resistive element 302 has the property that the resistance of the element changes with respect to the magnetic field caused by the current 306. This circuit requires the use of special technology which raises the cost of the device. Additionally, accuracy issues arise even though the current may be sensed with very low loss.

[1005] Referring now to Fig. 4, an alternative prior art technique for detecting current, through the use of a current transformer 402, is illustrated. As is shown, the current transformer 402 has a primary side 404 with a single loop and a secondary side 406 with multiple loops. A load resistance 408 is in parallel with the secondary side 406 of the transformer 402. The current transformer 402 is used to detect a current 410. The transformer 402 creates an output current equal to Ip/n, with I_p being the detected current and n being the turns ratio of the transformer 402. In this configuration, the resistance of the secondary side of the transformer is reflected to the primary side with the ratio 1/n². While current transformers work well for detecting currents, they are large and have a medium loss level and only work with alternating current (AC) circuits.

[1006] Another method for measuring currents involves the use of a Rogowski coil. Unfortunately, the voltage induced in a Rogowski coil is very small and easily disturbed when a measured current is less than, for example, 100 Amps. However, a Rogowski current transducer has a number of advantages over the current transformer illustrated in Fig. 4. For example, the Rogowski current sensor is linear, has no core saturation effects, has a wide bandwidth and a wide measurement range and is a relatively simple structure. The Rogowski coil comprises a toroidal winding placed around a conductor being measured. The Rogowski coil is effectively a mutual inductor coupled to the inductor being measured, where the output from the winding is an EMF proportional to the rate of change of current. While the above described techniques provide an indication of a sensed current in certain applications, the techniques, as noted above, individually have a number of short comings.

[1007] What is needed is a technique for detecting a current, within, for example, a power electronic circuit, that addresses many of the shortcomings of the prior art techniques described above.

SUMMARY

[1008] According to one embodiment of the present invention, a method is provided for generating a first current in at least one inductor in an integrated circuit die that is inductively coupled to a conductor carrying a second current. The conductor is disposed in an integrated circuit package holding the integrated circuit die. The method further includes integrating the first current in an integrator circuit on the die to provide a sensed voltage indicative of the second current.

[10091 In another embodiment, an apparatus is provided for sensing current that includes at least one inductor for generating a sensed current responsive to a current in a conductor when the inductor is inductively coupled to the conductor. The apparatus further includes an integrated circuit die including an integrator circuit coupled to the at least one inductor. The integrator circuit integrate!, the sensed current and generates a sensed voltage responsive to the sensed current, the sensed voltage indicative of the current in the conductor.

BRIEF DESCRIPTION OF THE DRAWINGS

[1010] The present invention may be better understood, and its numerous objects, features, and advantages made apparent to those skilled in the art by referencing the accompanying drawings.

[1011] Fig. 1 illustrates a prior art current sensor.

[1012] Fig. 2 illustrates a further prior art current sensor.

[1013] Fig. 3 illustrates yet another prior art current sensor.

[1014] Fig. 4 illustrates a further prior art current sensor.

[1015] Fig. 5a illustrates a coil in close proximity with a large current carrying wire, according to an embodiment of the present invention.

[1016] Fig. 5b illustrates a perspective cut away view of an integrated circuit, including a coupled coil and wire.

[1017] Fig. 6 is a cross-sectional view of a first embodiment of an integrated current sensor package.

[1018] Fig. 7 is a top view of the first embodiment of the integrated current sensor package. {1019] Fig. 8 is a model (equivalent circuit) of the integrated current sensor illustrated in Figs. 6 and 7.

[1020] Fig. 9 is a cross-sectional view of an alternative embodiment of the integrated current sensor package.

[1021] Fig. 10 is a cross section of the integrated current sensor package of Fig. 9 after the reflow processing step.

[1022] Fig. 11 is a model (equivalent circuit) of the alternative embodiment of the integrated current sensor package illustrated in Figs. 9 and 10.

[1023] Fig. 12 is a schematic diagram of an integrated current sensor.

[1024] Fig. 13 is a schematic diagram illustrating an integrated current sensor within a switched power supply circuit.

[1025] Fig. 14 is a timing diagram illustrating operation of the switched power supply circuit of Fig. 13.

[1026] Fig. 15 illustrates a further circuit for controlling a reset switch of an integrated current sensor.

[1027] Fig. 16 is a top view of a further embodiment of the integrated current sensor package.

[1028] Fig. 17 is a cross-sectional view of the embodiment of the integrated current sensor package in Fig. 16, along the line 17-17.

[1029] Fig. 18 is a further cross-sectional view of the embodiment of the integrated current sensor package in Fig. 16, along the line 18-18.

[1030] Fig. 19 is an electrical schematic of a circuit that includes a current sensor implemented in a buck converter application.

[1031] Fig. 19a is a timing diagram illustrating operation of the circuit of Fig. 19.

[1032] Fig. 20 illustrates a circuit that includes a current sensor with a reset circuit, configured according to one aspect of the present invention. [1033] Figs. 20a and 20b are timing diagrams that illustrate operation of the reset circuit of Fig. 20.

[1034] Figs. 21a illustrates an embodiment in which two coils are used to sense the current instead of one.

[1035] Fig. 21b illustrates another two coil embodiment in which the current carrying wire implemented on the current sensor package is "U" shaped.

[1036] Fig. 22 illustrates another view of an embodiment of a dual coil current sensor.

[1037] Fig. 23 A illustrates of how the dual coils are coupled to the integrator in an exemplary embodiment.

[1038] Fig. 23B illustrates a cross section of an exemplary dual coil embodiment.

[1039] Fig. 24 illustrates an embodiment in which the current carrying conductor is disposed on the printed circuit board.

[1040] Fig. 25 illustrates an embodiment in which the current carrying conductor is disposed on the substrate of the package carrying the integrated circuit die.

[1041] Fig. 26 illustrates an embodiment in which a four layer substrate is utilized and the inductors used to sense the current are disposed one a metal layer of the four layer substrate.

[1042] Fig. 27 illustrates a top view of the embodiment illustrated in Fig. 26.

[1043] Fig. 28 illustrates a top view of an embodiment in which a lead frame package is used.

[1044] Fig. 29 illustrates a cross sectional view of the embodiment illustrated in Fig. 28.

[1045] Fig. 30 illustrates another cross sectional view of the embodiment illustrated in Fig. 28.

[1046] Fig. 31 illustrates an embodiment in which the die is attached to a leadframe and the current carrying conductor carrying the current to be sensed is formed on the PCB [1047] Fig. 32 illustrates a top view of the embodiment shown in Fig. 31.

[1048] Fig. 33 illustrates a top view of the embodiment shown in Fig. 31.

[1049] Fig. 34 illustrates is cross sectional view of an embodiment in which a die is separated from the conductor by a glass dielectric on which the inductor(s) are patterned.

[1050] Fig. 35. illustrates a top view of the embodiment illustrated in Fig. 34.

[1051] Fig. 36 illustrates another cross sectional view of the embodiment illustrated in Figs. 34 and 35.

DESCRIPTION OF THE PREFERRED EMBODIMENTS

[1052] Referring now to the drawings, and more particularly to Fig. 5a, there is illustrated a coil or inductor 502 in close proximity with a large current carrying wire (or conductor) 504 such that the coil 502 and current carrying wire 504 act as coupled inductors. The coupled inductors, along with on-chip electronics, which will be discussed herein below, allow for the creation of the VSENSE signal which is proportional to an input current i_p in a manner that has very low loss, is very small and is a low cost implementation. This generally provides a better solution than the implementations described with respect to Figs. 1-4. The current provided through the current carrying wire 504 may be up to, for example, 1OA. Other embodiments may provide for more or less current carrying capability through the conductor. The coil 502 is placed near the current carrying wire 504 in order to create inductive coupling between the wire 504 and coil 502. As shown, the wire 504 overlaps only one side of the coil 502 such that the windings are all going the same way and the magnetic flux adds together. This causes an induced current in the other side of the coil 502 that is not overlapped by the wire 504.

[1053] Referring now to Fig. 5b, there is illustrated a perspective cut away view of the coil 502 and wire 504 illustrated in Fig. .5a. In this configuration one of the coupled inductors is placed within a silicon dioxide layer 604 on top of a die portion 606 of an integrated circuit chip. The coil 502 consists of metal runs in a metal layer, for example, an M5 layer, located within the silicon dioxide layer 604. The wire 504 would rest on the silicon dioxide layer 604 in close enough proximity to the coil 502, such that current passing through the wire 504 would induce another current within the portion of the coil 502 over which the wire 504 was not located.

[1054] There are multiple ways for implementing the coupled inductor configuration within a chip package. The first of these comprises a conductor 602, as illustrated in Fig. 6. The conductor 602 may be deposited on top of a silicon dioxide layer 604 of a die 606. The conductor 602 may comprise, for example, 15 μm of copper, or other suitable high conductivity material. As is illustrated, the inductor 502 is embedded within the silicon dioxide layer 604.

[1055] Referring now to Fig. 7, there is illustrated a top view of the package configuration. The conductor 602 is placed upon the silicon dioxide layer 604 of the die 606 (not shown in Fig. 7). The inductor 502 is located within the silicon dioxide layer 604 parallel to the wire 602. Bond wires 702 connect the conductor 602 on the die 606 with external outputs. The bond wires 702 typically support a maximum current of 1-2 amps, thus many bond wires are required to be connected to the conductor 602 for higher currents. Additional bond wires 704 connect portions of the die 606 to external pins 706 of the chip. Using the above-described package configuration, a 1OA sensor may be readily constructed.

[1056] Referring now to Fig. 8, there is provided a model (equivalent circuit) of the inductive coil package illustrated in Figs. 6 and 7. The coil 802 on the primary side- comprises a 500 pH coil. The coil 804 on the secondary side comprises a 2 μH coil. ^% Connected to a first side of the 500 pH coil 802 is a 1.5 mOhm resistor 808 in series with a 0.5 mOhm resistor 810. Connected to one output side of the 2 μH coil 804 is a 20 kOhm resistor 812. The 0.5 mOhm resistor 810 comprises the resistance provided by the coil 802. Since the copper wire 602 is not too thick and lies very close to the coil 502 of the chip, coupling coefficients between the copper wire 602 and the coil 502 are very good, assuming there is a distance of approximately two microns from the coil 502 (e.g., formed in an M5 layer) to the copper wire 602.

[1057] Referring now to Fig. 9, there is illustrated an alternative configuration wherein a package lead frame and flip chip configuration are used. A lead frame may be designed as follows. The die 902 is placed upside down with the silicon dioxide layer 904 suspended a short distance above a large copper slug or other suitable conductor 906. The copper slug 906 may have a large cross-sectional area for low loss. In this embodiment the slug 906 has a 200 x 200 μm cross-section. The die 902 is suspended above the conductor 906 on solder bumps 908, which rest on top of a lead frame 910. During device assembly, when heat is applied, the solder bumps 908 reflow causing the silicon dioxide layer 904 of the die 902 to come to rest directly upon the conductor 906.

[1058] Referring now to Fig. 10, there is illustrated a view of the embodiment of Fig. 9 with the silicon dioxide layer 904 of the die 902 resting on top of the conductor 906 after the reflow operation.

[1059] Referring now to Fig. 11, there is illustrated a circuit representation of the embodiment illustrated in Figs. 9 and 10. In this representation, a 200 x 200 μm copper slug 906 is utilized that is 3μm away from the coil 502, and the primary side includes a 520 pH coil 1102 in series with a 0.5 mOhm resistor 1104. The secondary side consists of 2 μH coil 1106 in series with a 20 kOhm resistor 1108. The coupling coefficient is reduced due to the lower current density in the slug.

[1060] With reference to Fig 16 there is illustrated a top view of a further configuration wherein a lead-on-chip (LOC) configuration is used. The die 1604 is connected to the lead frame 1602 by bond wires 1606. The die 1604 is attached to the current carrying conductor (wire) 1608 by, for example, double sided tape, non-conductive epoxy, or other dielectric . 1702. The current carrying conductor 1608 is inductively coupled to an inductor in the die 1604. Note that although the dielectric 1702 is illustrated as wider than the conductor for ease of understanding, in preferred embodiment, the dielectric 1702, is the same width as the conductor.

[1061] Referring now to Fig. 17, there is illustrated a cross-sectional view of Fig. 16, along line 17-17. The die 1604 is connected to the conductor 1608, via the tape 1702, as described previously. The die 1604 connects to the lead frame 1602, via bond wires 1606. The tape 1702 may be, for example, approximately 75 μm thick. As is shown, the structure is contained within a mold compound 1704. Referring now to Fig. 18, there is illustrated a cross sectional view of Fig. 16, along line 18-18.

[1062] Turning to Fig. 12, there is illustrated a schematic diagram of the electronic circuit necessary for recreating the VSENSE signal when detecting the current i_p using the coupled inductors as illustrated in Figs. 5 A and 5B. The coupled inductor 1202 comprises either of the configuration packages described hereinabove or, alternatively, may comprise a different undescribed configuration package that places the coil in close proximity with the wire to inductively couple them together. The primary side is modeled by inductor 1204 in series with resistor 1206. The secondary side is modeled by inductor 1208 which is connected to a resistor 1210. The resistor 1210 is then connected to ground. The other side of inductor 1208, which is connected to the negative (inverting) input of an operational amplifier 1212, outputs the induced current I_n. The positive (non-inverting) input of operational amplifier 1212 is connected to ground.

[1063] The current through the secondary is dominated by the resistive loss of resistor 1210 and is the derivative of the primary current. An integrator circuit 1218 is used to integrate the induced current I_n- The integrator circuit 1218 includes the operational amplifier 1212, a capacitor 1214 (connected between the output of operational amplifier 1212 and the negative input of operational amplifier 1212) and a reset switch 1216 (connected between the output of operational amplifier 1212 and the negative input of operational amplifier 1212) in parallel wxth the capacitor 1214. Thus, the current I_n may be determined according to the equation:

I_n = (WRiXdip/dt)

By integrating on the capacitor 1214 an output voltage V_SENSE is attained according to the following equation:

VS_ENS_E = 1/C J Indt = (LnTRiC)ip

In this case, L_n,, the mutual inductance, is well controlled, but can vary from part to part due to assembly variations. The capacitance C will vary from part to part and probably can be controlled to +/- 5% accuracy. The capacitor 1214 will generally not have any appreciable temperature coefficient. R₁ is dominated by the metal resistance of the coil and will vary from part to part and is equal to the value of the resistor 1210 and also has a large temperature coefficient.

[1064] In order to obtain overall accuracy for the capacitance C which varies from part to part, factory calibration using a one time programmable (OTP) memory 1220 can be used. In a preferred embodiment, a low cost 32-bit OTP memory may be utilized. The OTP memory 1220 provides a control variable to a programmable gain amplifier 1222. The first gain stage 1223, consisting of programmable amplifier 1222, programmable resistance 1224 and the OTP memory 1220, compensates for part to part variations of the circuit. The OTP memory 1220 is programmed at the factory based upon measurements made there. The programmable gain amplifier 1222 has its negative input connected to the output of the operational amplifier 1212. A programmable resistance 1224 is connected between the output of the programmable amplifier 1222 and ground. The positive input of programmable amplifier 1222 is connected to the programmable resistance 1224. The value of the programmable resistance 1224, and thus the gain of the first gain stage 1223, is controlled by the values provided from the OTP memory 1220.

[1065] A second gain stage 1226 compensates for differences in the resistance caused by temperature variations in the device. A temperature sensor 1228 and an analog-to-digital converter (ADC) 1230 are used to generate a digital temperature value to compensate for the coil resistance temperature coefficient. The temperature sensor 1228 detects the temperature and generates an analog repiesentation of the temperature. The ADC 1230 converts the analog signal into a digital signal. The digital temperature value is provided, via a control bus 1231, to control logic 1232. In one embodiment the control logic 1232 may consist of a look-up table. The look -up table would include various control values associated with particular temperature values. Alternative embodiments may include a microprocessor programmed to control the output according to various temperature levels or other types of digital logic. The control logic 1232 provides a control value to the programmable gain amplifier 1234 and programmable resistance 1236. The negative input of the amplifier 1234 is connected to the output of programmable amplifier 1222. The programmable resistor 1236 is connected between the output of programmable amplifier 1234 and ground. The positive input of ihe amplifier 1234 is connected to the programmable resistance 1236. The particular value of the programmable resistance 1236, and thus the gain of the second gain stage 1226, is controlled via the output from the control logic 1232. The output of the amplifier 1234 provides the compensated V_SENSE signal. The code provided by the control logic 1232 is updated during the phase in which the operational amplifier 1212 is reset responsive to a reset signal applied to switch 1216. The reset signal is applied while the sensed current i_p is zero.

[1066] The current sensor is designed to be used in, for example, a switched power supply. When the current i_p is equal to zero, a reset signal maybe applied to switch 1216 to reset the capacitor 1214, and the logic value applied to amplifier 1234, via control logic 1232, is updated responsive to the presently sensed temperature from temperature sensor 1228. Referring now to Fig. 13, there is provided one example of how to apply the reset signal to a current sensor 1302 within a buck converter circuit. In this case, the buck converter circuit control signal φ₂ is applied to a transistor 1304 having its drain/source path connected between 12 volts and node 1306. A second transistor 1308 has its drain/source path connected between node 1306 and node 1310. The transistor 1308 is controlled by a second control signal φi. The current sensor 1302 is connected between node 1310 and ground to detect current i_p and provide a control signal V_SENSE- An inductor 1312 is connected between node 1306 and node 1314. A capacitor 1316 is connected between node 1314 and ground. A load 1318 is also connected between node 1314 and ground. In one embodiment, the reset signal to switch 1216 of the current sensor 1302 may be configured to be the control signal φ₂.

[1067] As illustrated in Fig. 14, the current i_p is zero when signal φi goes low and when signal φ₂ goes high at, for example, time.ti. Integrator 1218 is reset during phase two when signal φ₂ goes high and the current sensor would accept signal φ₂ as an input to drive the reset signal to switch 1216, since the current i_p is zero during this time. As can be seen each time the signal φ₂ goes high, the current i_p is zero enabling the reset signal to be applied to the integrator circuit 1218.

[1068] Referring now to Fig. 15, there is illustrated an alternative embodiment wherein the reset signal to the reset switch 1216 is generated responsive to a one-shot circuit consisting of negative glitch detect circuit 1502 and one-shot circuit 1504. When the current i_p goes low as illustrated, for example, at ti in Fig. 14, the negative glitch detect circuit 1502 will detect the negative edge of current i_p. In response to this detection, the negative glitch detect circuit 1502 generates a pulse to the one-shot circuit 1504. The one- shot circuit 1504 then generates the reset signal to the reset switch 1216 responsive to the pulse from the negative glitch detect circuit 1502. Other methods for detecting when the sensed current i_p goes to zero may also be utilized for generating the reset signal to reset switch 1216. The examples illustrated in Figs. 13-15 are merely provided as examples of some embodiments thereof. [1069] With reference to Fig. 19, a relevant portion of a buck power converter 1900 is depicted that includes a current sensor 1902 that senses a current provided by a power source H-V. Referring to Fig. 19a, a timing diagram 1970 depicts various waveforms used in conjunction with the converter 1900. In this application, a signal φi drives switch 1904 to cause a current 'i' to flow (from the power source +V through an inductor 1920) and power to be provided to a load 1912, through an inductor Ll and a capacitor Cl. The current 'i' that flows through the inductor 1920 induces a current that flows through the inductor 1922 and resistor 1924, charging capacitor 1926. When the switch 1904 is conducting, switch 1906 is in a non-conducting state. Similarly, when the switch 1904 is in a non-conducting state, a signal φ₂ drives switch 1906 into a conducting state. The switches 1904 and 1906 may be, for example, implemented as enhanqement-mode field-effect transistors (FETs).

[1070] As is also shown, the signal ψ₂ may also function as a reset signal to drive a switch 1908 which, when conducting, shorts the capacitor 1926 of integrator 1910 and resets the integrator 1910 of the current uensor 1902. It should be appreciated that it is desirable for the reset signal to be turned off prior to a time when the current 'i' again flows through inductor 1920 in order to not adversely affect an output signal VOUT provided by the integrator 1910. In modern power converters this requirement can be difficult to meet as the falling edge of the signal φ₂ and the rising edge of the signal φi may slightly overlap. In this event, the integrator 1910 may still be in a reset state when the current T again begins flowing through the inductor 1920. As a result, the output signal VOUT (provided by the integrator 1910) will not provide an accurate indication of the current T that flows through the inductor 1920. This is particularly true for a current, such as the current T shown in Fig. 19a, where the induced current is proportional to di/dt and the change in 'i' is most pronounced at the start of φi.

[1071] Turning to Fig. 20, a reset ciircuit 1950, configured according to one aspect of the present invention, is shown implemented in conjunction with the current sensor 1902 of Fig. 19. A control signal, as is shown in timing diagrams 1980 and 1990 of Figs. 20a and 20b, respectively, corresponds to the signal φ₂ of Fig. 19a. It should, however, be appreciated that the control signal may be a different signal having, for example, a different polarity. The reset circuit 1950 includes a one-shot multivibrator 1960 and an AND gate (or one or more logic gates) 1962. As is used herein, the term "one-shot circuit" or "one-shot multivibrator" is a device with one stable state that, responsive to an input signal, provides an output signal for a period of time before returning to the stable state. A reset signal is provided by the AND gate 1962, responsive to the control signal and a reset_one signal provided at the output of the multivibrator 1960. A pulse width of the reset signal corresponds to a pulse width of the shortest one of the reset and reset_one signals. As such, except for the case of extremely short pulse widths of control signals φi and (J)₂, depending upon the implementation, the reset circuit 1950 ensures that the reset of the integrator is shorter than the control signals and thus the integrator reset is released before the current rises substantially above zero, thus allowing accurate sensing the current. It should be appreciated that the above-described reset circuit can be readily incorporated within the same integrated circuit that includes the current sensor and the use of φ₂ (or other signal) readily allows for external timing control. Thus, one of the external pins 706 (Fig. 7) can be supplied with an appropriate φt or φ₂ control signal for use to generate the reset signal, depending upon the location of the current sensor. Exemplary placements of the current sensor for a buck converter are illustrated in Figs. 13 and 19.

[1072] In another embodiment, two coils are used instead of one to sense the current. Referring to Fig. 21a a top view of two coils 2101 and 2103 are seen in die 2105 vertically displaced from the current carrying wire 2107. The coils 2101 and 2103, shown as single turns for ease of illustration, may be mulliple turn coils. In an exemplary embodiment the width of the current carrying conductor is; 0.75mm. The use of two inductors allows significant cancellation of stray fields from external sources, and thus more accurate current sensing in certain environments. In addition, use of two inductors allows more current to be induced, e.g., twice as much current, as compared to single inductor embodiments. In another embodiment illustrated in Fig. 21b, the current carrying conductor 2109 has a different configuration and is shaped as a "U". Potential interferers, conductors 2111 and 2115, are also illustrated. In both embodiments, the current carrying conductors 2107 and 2109 may be copper and formed as part of the lead frame used in the package housing the die with the dual coils. Fig 22 illustrates another view of an embodiment of a dual coil current sensor implementation having two coils 2201 and 2203 on die 2205 separated from current carrying conductor 2207 by a dielectric interposer 2209. The dielectric interposer may be glass, ceramic, oxidized silicon, B-stage epoxy, double stick tape, or other suitable dielectric material. [1073] Fig. 23 A illustrates an exemplary embodiment of how the dual coils are coupled to the integrator. For ease of illustration, the reset circuit and the calibration and temperature compensation circuits are omitted from the figure. As shown in Figs. 23 A and 23B , the direction of current in coil 2301; and 2303 is caused by the magnetic flux generated by the current flowing conductor 2207. As shown in Fig. 23 A the direction of the flux generated by the current flowing conductor 2207 is in the opposite direction for the two coils (one into the page and one out of the page), thus resulting in a substantial improvement in cancellation of interference from far field interferers over a single coil embodiment in certain applications and increased induced current from current in conductor 2207. Note also that in Fig. 23 single turn coils are illustrated. In other embodiments the coils may be implemented with the appropriate number of turns to adequately provide the current sensing capability.

[1074] Fig. 23B illustrates a cross sectional view of the two inductor embodiment illustrated in Fig. 22. The magnetic flux 2208 from the current flowing in conductor 2207 is illustrated.

[1075] While the embodiments illustrated above generally show the current carrying conductor incorporated in the package carrying the integrated circuit die, other embodiments contemplate the current caitying conductor being provided on, e.g., the printed circuit board (PCB) on which th& current sensor is mounted. One such embodiment is illustrated in Fig. 24. Device 2400 includes die 2401 mounted on a two layer substrate 2403. The die includes inductors 2404 used in the current sensor and the sensing circuitry shown, e.g., in Figs. 12 and 20. The two layer substrate has a metal layer on surfaces 2405 and 2407. The substrate may be formed by any appropriate substrate material such as BT, FR4, flextape, ceramic, or other suitable material. The substrate provides a high voltage dielectric between the current carrying conductor 2409 and die 2401. The current carrying conductor 2409 is a trace formed on PCB 2410. Vias 2411 and bondwires 2415 provide connections between the die 2401 and additional traces 2413 on printed circuit board 2410. Metal pads 2412 on the package are connected via solder joint 2414 to traces 2413. The bondwires and die are encapsulated in a plastic mold 2417. Moving the current carrying conductor off of the package makes the package less expensive to manufacture but can result in more difficult calibration of the current sensor as it may need to be calibrated after mounting on the printed circuit board 2410 depending upon the accuracy needed. [1076] In another embodiment, illustrated in Fig. 25, the current carrying conductor 2501 is formed on the substrate 2503. In the embodiment illustrated in Fig. 24, when the package 2400 is soldered down to the PCB 2410, the gap 2416 between the device 2400 and the trace 2409 will vary slightly because of the variability in the height of the solder joints 2414 coupling device 2400 to printed circuit board 2410. That affects the accuracy of the sensing and may require calibration after the device is mounted on the board. However, in the embodiment illustrated in Fig. 25, the current carrying conductor is part of device 2500. That means the distance between the current carrying conductor 2501 and the inductors 2504 is defined solely by the dielectric 2503 and not by the variability of the height of the solder joints connecting the device 2500 to a PCB board. That makes the device more accurate and additionally allows the device 2500 to be calibrated in the factory prior to mounting on the PCB. In comparing Fig_; 25 and Figs. 16-18 note that a major difference is that the die is separated from the current carrying conductor by a dielectric such as tape in Figs. 16-18 and by a high voltage dielectric in Fig. 25. The embodiment shown in Fig. 25 can provide higher voltage isolation because there is greater separation because the dielectric is thicker and the dielectric can be a better quality dielectric.

[1077] Referring to Fig. 26, in another embodiment, the inductors 2601 used to sense the current in the current carrying conductor 2619 are formed on a metal layer of four layer substrate 2603. The die 2605 is connected to the coils via bond wires 2609 and vias 2611 formed in the substrate. Additional vias 2615 and bondwires 2617 provide external connections for the die 2605. Finally, the current carrying conductor 2619 is formed either on the PCB as in Fig. 24 or on the package substrate as in Fig. 25.

[1078] One advantage of the approach in Fig. 26 is that the current /carrying conductor can be made wider than the die. Fig. 27 illustrates a top view showing that the current carrying conductor is placed between the two coils but is significantly bigger than the die. That allows significantly more current to, be carried than in alternative embodiments in which the current carrying conductor is limited by the size of the die.

[1079] Referring now to Fig. 28, illustrated is a top view of an embodiment in which a device 2800 is a lead frame package supporting die 2804 on a lead frame 2808. The current carrying conductor is provided by lead frame 2808. In an embodiment, the lead frame 2808 is copper. The lead frame/conductor 2808 is separated from the die 2804 by dielectric interposer 2802. Bond wires 2806 provide connections between the die 2801 and external connections 2809 of the package.

[1080] Referring now to Fig. 29, there is illustrated a cross sectional view of Fig. 28, along line 29-29. Note that the dielectric interposer 2802 may be, e.g., glass, ceramic, oxidized silicon, B-stage epoxy, double stick tape, or other suitable dielectric material. Fig. 30 illustrates a cross sectional view of Fig. 28, along line 30-30.

[1081] Fig. 31 illustrates another embodiment in which the die 3101 is mounted on a leadframe 3102. However, the current carrying conductor 3105 carrying the current to be sensed is formed on the PCB rather than using the leadframe paddle 3101, which supports die 3101. Fig. 32 illustrates a top view of the embodiment shown in Fig. 31. Fig. 33 illustrates another top view of the device shown in Fig. 31 illustrating the current carrying conductor 3105 under the die 3101.

[1082] Referring now to Fig. 34, illustrated is cross sectional view of an embodiment in which a die 3401 is separated from the conductor 3404, e.g., formed by the leadframe, by a glass dielectric 3405. However, unlike the embodiment, in Fig. 30, the inductor is patterned on the glass dielectric rather than formed in metal layers in the die 3401. That allows the die to be shrunk as the inductors may be a factor determining a minimum size of the die. Forming the inductor(s) on top of the glass dielectric may be done in manner well known in the art, e.g., using copper, appropriate mask, and photoresist. Fig. 35 shows a top view of the embodiment illustrated in Fig. 34. Fig. 36 shows another cross sectional view of the embodiment illustrated in Fig. 35. After the inductors are formed, a passivation layer 3602 may be formed on the inductors. Note that because the inductors are formed on the glass dielectric, the inductors may be larger than those formed in the die and thus more sensitive. The glass dielectric with a thickness of 100 microns provides, e.g., five to ten thousand volts isolation between the die 3401 and the current carrying conductor.

[1083] Although various embodiments have been described in detail, it should be understood that various changes, substitutions and alterations can be made therein without departing from the scope of the invention, as defined by the appended claims.

Claims

WHAT IS CLAIMED IS:

1. An apparatus for sensing current, comprising: at least one inductor for generating a sensed current responsive to a current in a conductor when the inductor is inductively coupled to the conductor; and an integrated circuit die including an integrator circuit coupled to the at least one inductor for generating a sensed voltage responsive to the sensed current, the sensed voltage indicative of the current in the conductor.

2. The apparatus for sensing current as recited in claim 1 wherein the at least one inductor is formed in the integrator circuit die.

3. The apparatus for sensing current as recited in claim 1 wherein the at least one inductor is formed in an integrated circuit package containing an integrated circuit die.

4. The apparatus for sensing current as recited in any of claims 1 to 3 further comprising an integrated circuit package containing the integrated circuit die and the conductor is formed within of the integrated circuit package.

5. The apparatus for sensing current as recited in any of claims 1 to 3 further comprising an integrated circuit package containing an integrated circuit die wherein the conductor is formed as part of a printed circuit board on which the integrated circuit package is mounted.

6. The apparatus for sensing current as recited in claim 4, further including a lead frame for supporting the integrated circuit die.

7. The apparatus for sensing current as recited in any of claims 1 to 5, wherein the integrator circuit further comprises: an operational amplifier for generating the sensed voltage responsive to the output current; a capacitor connected between an input of the operational amplifier and the output of the operational amplifier; and a switch connected between the input of the operation amplifier and the output of the operational amplifier for resetting the integrator circuit responsive to a reset signal.

8. The apparatus for sensing current as recited in any of claims 1 to 7, further including: a first compensation circuit for compensating the sensed voltage for part differences in the integrator circuit; and a second compensation circuit for compensating the sensed voltage responsive to a sensed temperature.

9. The apparatus for sensing current as recited in claim 8, wherein the first compensation circuit further comprises: a first programmable amplifier for compensating the sensed voltage responsive to a first control value , the first control value configuring the first programmable amplifier to compensate for part to part differences in the integrator circuit; and a memory for storing the iϊrst control value; and wherein the second compensation circuit further comprises: a second programmable amplifier for compensating the sensed voltage responsive to a second control value, the second control value configuring the second programmable amplifier to compensate for temperature differences; a temperature sensor for generating a temperature signal; and control logic responsive to the temperature signal for generating the second control value.

10. The apparatus for sensing current as recited in any of claims 1 to 9, wherein the integrator circuit is reset responsive to a reset control signal.

11. The apparatus for sensing current as recited in claim 10, wherein the reset control signal is determined according to a switching control signal for a switched power supply.

12. The apparatus as recited in any of claims 1 to 11 wherein the one or more inductors comprise two inductors coupled to the integrator circuit.

13. The apparatus as recited in any of claims 1 to 12, further including a lead frame for supporting the integrated circuit die and for carrying the current to be sensed.

14. The apparatus as recited any of claims 1 to 13 further comprising a glass dielectric disposed between the current carrying conductor and the die, the one or more inductors being formed on the glass dielectric.

15. A method comprising: generating a first current in at least one inductor in an integrated circuit die that is inductively coupled to a conductor carrying a second current, the conductor in an integrated circuit package holding the integrated circuit die; and integrating the first current in an integrator circuit on the die to provide a sensed voltage indicative of the current.

16. The method as recited in claim 15 further comprising resetting the integrator circuit with a reset control signal determined according to a switching control signal for a switched power supply.

17. The method as recited in any of claims 15 to 16 further compensating the sensed voltage for part differences in the integral or circuit.

18. The method as recited in any of claims 15 to 17 further comprising compensating the sensed voltage responsive to a sensed temperature.

19. The method as recited in any of claims 15 to 18 further comprising sensing the first current in two inductors on the integrated circuit and supplying the current induced in the two inductors to the integrator circuit.