Hall-effect-based magnetic field sensor having an improved output bandwidth
Granted 30 May 2017 · 1 office action
Assignee: STMicroelectronics
Law firm: Law firm · Log in to unlock
Attorney: Attorney · Log in to unlock
Inventors: Marco Marchesi, Giulio Ricotti, Marco Morelli · Examiner: Patrick Assouad · AU 2858 · TC 2800
Life of the application
9 dated eventsAbstract
A magnetic field sensor formed by a Hall cell having a first, second, third and fourth conduction nodes electrically coupled together by resistive paths. Flowing between the first and second conduction nodes is a control current. In the presence of a magnetic field, a difference of potential due to the Hall effect is generated between the third and fourth conduction nodes. An operational amplifier has an inverting input terminal coupled to the fourth conduction node, a non-inverting input terminal biased at the voltage at the third conduction node, and an output terminal coupled in feedback mode to the inverting input by a feedback resistor. The current generated in feedback through the feedback resistor generates a voltage indicating unbalancing, due to the Hall effect, between the third and fourth conductive nodes, and consequently indicates the intensity of the magnetic field that acts upon the Hall cell.
Description
10 parts›PRIORITY CLAIM
This application claims priority from Italian Application for Patent No. TO2014A000261 filed Mar. 27, 2014, the disclosure of which is incorporated by reference.
›TECHNICAL FIELD
The present invention relates to a Hall-effect-based magnetic field sensor.
›BACKGROUND
As is known, is frequently needed to measure magnetic fields accurately. For this purpose, various solutions have been proposed. In particular, there has already been proposed use of Hall sensors that exploit the well-known Hall effect, whereby, if a current flows through a conductor immersed in a magnetic field, the latter exerts a force transverse on the charge carriers that flow in the conductor and said force tends to “push” said charge carriers onto one side of the conductor. This effect is particularly evident in a thin flat conductor. The accumulation of the charges on the sides of the conductor determines a measurable voltage between the sides of the conductor itself and thus represents a measurement of the magnetic field.
A Hall sensor may be represented schematically as a resistive Wheatstone bridge, as illustrated in FIG. 1A , and is fabricated by forming resistive wells, for example implanted in a semiconductor or insulating substrate, accessible from outside by conductive pads P 1 -P 4 , as illustrated schematically, in top plan view, in FIG. 1B . The die of FIG. 1B provides a diffused resistive matrix, in which the resistive wells associated to two inputs I 1 and I 2 are electrically coupled to the resistive wells associated to two outputs O 1 and O 2 , for example according to the connection scheme of FIG. 1A . Between the inputs I 1 and I 2 there flows, in use, a biasing, or control, current i BIAS . In the presence of a magnetic field B in which the resistive bridge of FIGS. 1A, 1B is immersed, the output voltage V HALL , measured as electrical potential difference that is set up across the outputs O 1 and O 2 , assumes a value different from (greater in modulus than) the value that may be measured in resting conditions, i.e., in the absence of magnetic field B. Said variation is a function of the value of the magnetic field B in which the Hall sensor is immersed (the voltage V HALL is proportional to the magnetic field B).
Reading of the voltage V HALL enables acquisition of information on the magnetic field B (field intensity and direction). However, there may exist different elements that tend to limit the frequency response of the Hall sensor. For instance, the parasitic capacitances associated to the output nodes O 1 and O 2 , which vary as a function of the dimensions of the Hall sensor, may limit the operating bandwidth to a few tens of kilohertz, which is insufficient for some types of applications (for example, the driver circuits for control of electric motors).
There is accordingly a need to provide a magnetic field sensor exploiting the Hall effect that is such as to overcome the drawbacks of the known art.
›SUMMARY
In an embodiment, a magnetic field sensor comprises a first Hall cell having a first conduction node, a second conduction node, a third conduction node and a fourth conduction node electrically coupled together by one or more resistive paths. The first and second conduction nodes are biasing inputs of the first Hall cell, which can be biased for generating a flow of control current through said one or more resistive paths. The third and fourth conduction nodes are outputs of the first Hall cell, wherein, in an operating condition of presence of a magnetic field acting upon the first Hall cell, a Hall voltage is generated between said third conduction node and said fourth conduction node. A first biasing circuit is operatively coupled between said first and third conduction nodes and configured to bias the third conduction node at a first operating voltage and to bias the first conduction node at a second operating voltage to generate said flow of control current. A second biasing circuit has a first input terminal operatively coupled to the fourth conduction node, a second input terminal adapted to receive a third operating voltage that is a function of the first operating voltage, and an output terminal coupled in a feedback mode to the first input by a detection module. The second biasing circuit is configured to generate, during said operating condition of presence of the magnetic field, a feedback current through the detection module such as to maintain the fourth conduction node at the voltage value defined by the third operating voltage. The detection module is configured to generate an output signal that is a function of said feedback current.
›BRIEF DESCRIPTION OF THE DRAWINGS
For a better understanding of the present invention, preferred embodiments thereof are now described, purely by way of non-limiting example and with reference to the attached drawings, wherein:
FIGS. 1A and 1B are a schematic illustration of a Hall cell of a known type, according to a circuit representation based upon a resistive Wheatstone bridge and in top plan view of a die that integrates the Wheatstone bridge;
FIG. 2 shows, in circuit form, a magnetic field sensor according to one embodiment;
FIG. 3 shows the magnetic field sensor of FIG. 2 including a current-spinning circuit, for compensation of non-idealities; and
FIG. 4 is a circuit representation of a magnetic field sensor according to a further embodiment.
›DETAILED DESCRIPTION OF THE DRAWINGS · 1 of 5
According to one aspect a magnetic field sensor is provided, comprising a Hall cell having a first conduction node I 1 , a second conduction node I 2 , a third conduction node O 1 , and a fourth conduction node O 2 electrically coupled together by one or more resistive paths R 1 -R 4 . In particular, said resistive paths R 1 -R 4 provide a Wheatstone bridge (see, for example, FIGS. 1A and 1B ). The first and second conduction nodes are, respectively, a first biasing input I 1 and a second biasing input I 2 of the Hall cell, flowing between which is a biasing current i BIAS , as is known. The third and fourth conduction nodes are, respectively, a first output O 1 and a second output O 2 of the Hall cell, across which a Hall voltage is set up in the presence of an external magnetic field B acting upon the Hall cell. According to one embodiment (for example, as shown in FIG. 2 ) further present is a first operational amplifier (designated by the reference 20 ) having its inverting input coupled to the first output O 1 of the Hall cell, its non-inverting input biased at a first operating voltage, and an output terminal coupled to the first biasing input I 1 of the Hall cell (and thus fed back onto the inverting input via the resistive path R 1 of the Wheatstone bridge). The second biasing input I 2 of the Hall cell provides a reference node, for example at ground voltage.
For operation of the first operational amplifier, the voltage set at the non-inverting input is (ideally) brought to the inverting input, which, in this case, corresponds to the first output O 1 of the Hall cell. In this way, the first output O 1 of the Hall cell is set at virtual ground, and the parasitic capacitance associated to said conduction node O 1 is not charged/discharged significantly in use.
Once again with reference to the embodiment of FIG. 2 , a second operational amplifier is further present (designated by the reference number 22 ), having an inverting input terminal coupled to the second output O 2 of the Hall cell and a non-inverting input terminal biased at a voltage that is a function of (for example, equal to) the voltage present on the first output O 1 of the Hall cell. The output terminal of the second operational amplifier is feedback-connected to the inverting input of the same second operational amplifier by a feedback resistor. The feedback current generated through the feedback resistor generates, across the latter, a voltage that is dependent upon the Hall-effect unbalancing between the first and second outputs O 1 and O 2 and consequently indicates the intensity of the magnetic field B that acts upon the Hall cell. As may well be understood, according to this embodiment of the present invention, both the first output O 1 and the second output O 2 of the Hall cell are at virtual ground, as a result of operation of the operational amplifiers 20 and 22 . The current generated at output from the second operational amplifier consequently flows through the feedback resistance and is again injected into the Hall cell to compensate the unbalancing generated by the Hall voltage in the presence of magnetic field. It is consequently evident that the current that traverses the feedback resistor, which may be detected as voltage across the feedback resistor itself, indicates the Hall potential V HALL between the outputs O 1 and O 2 of the Hall cell. The potential V HALL in turn indicates the intensity of the magnetic field B to be measured. The relation that exists between the intensity of the magnetic field B to be measured and the value of voltage across the feedback resistor may, for example, be detected experimentally by tests in conditions of magnetic field of variable intensity. For a plurality of known values of intensity of the magnetic field that acts upon the Hall cell a respective voltage value across the feedback resistor is acquired, and an experimental relationship that links the output of the magnetic field sensor to the magnetic field to be measured is thus obtained.
It is evident that the first and second operational amplifiers may be replaced by circuit elements of another type, which are designed to perform the same functions, or similar functions, i.e., bias the outputs of the Hall cell to prevent fluctuations of the charge of the parasitic capacitance at said conduction nodes, generating the biasing current between the inputs of the Hall cell, and generating an output current, the value of which is a function of the Hall voltage resulting from an external magnetic field. Likewise, the feedback resistor may be obtained according to any technology, comprising electrical/electronic elements of a passive and/or active type.
Preventing, or significantly reducing, phenomena of charge and discharge of the parasitic capacitances associated to the outputs of the Hall cell has as positive effect increase of the operating bandwidth of the magnetic field sensor. Generation of an output signal (indicating the magnetic field) in the form of current signal presents considerable advantages and in particular enables mutual connection in parallel of a plurality of Hall cells and makes it possible to add together all the output-current signals from each of said Hall cells to obtain an amplified signal that is easier to read.
There now follows a more detailed description of a preferred embodiment of the magnetic field sensor. In particular, a magnetic field sensor 1 , including a Hall cell 3 of the type shown in FIGS. 1A, 1B , is illustrated schematically in FIG. 2 , according to one aspect. The Hall cell 3 includes: a first conduction terminal 2 , here having the function of first input (I 1 ); a second conduction terminal 4 , here having the function of second input (I 2 ), coupled to a reference voltage, for example a ground reference voltage GND; a third conduction terminal 6 , here having the function of first output (O 1 ); and a fourth conduction terminal 8 , here having the function of second output (O 2 ). For instance, as is shown in FIG. 1A , each of the terminals 2 , 4 , 6 and 8 forms a conductive node of a resistive Wheatstone bridge. According to this configuration, the first input I 1 is electrically coupled to the first output O 1 by a resistive path (referred to hereinafter more simply as “resistor”) 10 , having a value of resistance R 1 ; the first output O 1 is electrically coupled to the second input I 2 by a respective resistive path (referred to hereinafter more simply as “resistor”) 12 , having a value of resistance R 2 ; the first input I 1 is electrically coupled to the second output O 2 by a resistive path (referred to hereinafter more simply as “resistor”) 14 , having a value of resistance R 3 ; and the second output O 2 is electrically coupled to the second input I 2 by a respective resistive path (referred to hereinafter more simply as “resistor”) 16 , having a value of resistance R 4 .
›DETAILED DESCRIPTION OF THE DRAWINGS · 2 of 5
It is, however, evident that the inputs I 1 , I 2 and outputs O 1 , O 2 of the Hall cell 3 may be electrically coupled together according to any configuration, and in particular the electrical connection between the inputs I 1 and I 2 and the outputs O 1 and O 2 may be different from the one shown. For instance, there may be, in addition or as an alternative to the electrical couplings shown in FIG. 2 , a further resistive connection between the first input I 1 and the second input I 2 , and/or a further resistive connection between the outputs O 1 and O 2 .
In general, in use, flowing between the input terminals I 1 , I 2 is a biasing current i BIAS . Generated between the output terminals O 1 , O 2 , in use and in the presence of a magnetic field B, is a potential difference due to the Hall effect. The input terminals and the output terminals are arranged to be aligned with respect to one another in respective substantially mutually orthogonal directions that define a plane that is parallel to the plane of lie of the Hall cell 3 . A magnetic field sensor 1 having such a Hall cell 3 is sensitive to components of a magnetic field B that are orthogonal to the plane of lie of the Hall cell 3 . See, for example, FIGS. 1A and 1B , which show a Hall cell of this type.
According to one aspect, the magnetic field sensor 1 further comprises a first operational amplifier 20 , supplied in voltage between −Vcc and +Vcc (e.g., between −15V and +15V), having its inverting input 20 a electrically coupled to the first output O 1 and its non-inverting input 20 b electrically biased to a predefined voltage value V force , in particular equal to Vdd/2. The voltage Vdd is chosen according to the need, as compromise between the maximum voltage available in the application considered and the sensitivity that it is desired to obtain for the Hall sensor 1 (in fact, the voltage at output from the Hall sensor is proportional to the supply voltage). For instance, the voltage Vdd is chosen in the range 2-15 V, more in particular 3-5 V, for example 4 V. The output 20 c of the operational amplifier 20 is electrically coupled to the first input I 1 , thus biasing it and generating the flow of the biasing current i BIAS between the nodes 2 and 4 . Thus, as may be noted, the output 20 c of the first operational amplifier 20 is electrically coupled to the first input 20 a of the first operational amplifier 20 through the resistor 10 .
The magnetic field sensor 1 further comprises a second operational amplifier 22 , supplied in voltage between −Vcc and +Vcc (e.g., between −15V and +15V), having its inverting input 22 a electrically coupled to the second output O 2 and its non-inverting input 22 b electrically biased at a predefined reference voltage V ref , for example equal to the voltage V O1 present on the first output O 1 . The output 22 c of the operational amplifier 22 is fed back onto the inverting input 22 a of the amplifier 22 itself through a feedback branch 24 that includes a feedback resistor 26 . The feedback resistor is, for example, a passive resistive element. However, the feedback resistor may be obtained according to any technology, and may be formed by, or comprise, active elements (e.g., transistors) and/or passive elements. For instance, the electrical resistance of the feedback resistor 26 is comprised between 50 and 300 kΩ, in particular between 100 and 200 kΩ, for example approximately 150 kΩ.
As may be noted, the first operational amplifier 20 , the resistor 10 , and the resistor 12 provide, together, a non-inverting-amplifier configuration for the first operational amplifier 20 . The input signal V force , applied to the input 20 b of the operational amplifier 20 , and the divider formed by the resistors 10 and 12 , provide a negative reaction. Since, ideally, the voltage applied to the input 20 b is equal to the voltage applied to the input 20 a of the operational amplifier 20 , we have that the voltage V O1 at the input 20 a is equal to the voltage V force at the input 20 b , and both are equal to Vdd/2. The signal V u , generated at output from the first operational amplifier 20 , is equal to V I1 =V force ·(R 1 +R 2 )/R 2 .
Introducing a further idealization, and assuming that the resistors 10 , 12 , 14 and 16 all have the same a resistance (i.e., R 1 =R 2 =R 3 =R 4 ), we have V I1 =Vdd. Consequently, also the voltage at the first input I 1 of the Hall cell 3 is Vdd. It is evident that, following the idealizations so far made, on each resistor 8 , 10 , 12 and 16 we have a voltage drop equal to Vdd/2. It follows that the voltage V O2 present to the output O 2 , set at input to the terminal 22 a of the second operational amplifier 22 , is also equal to Vdd/2.
In the absence of an external magnetic field B, it is desirable for the output of the magnetic field sensor 1 to be equal to a predefined value, for example to the voltage at input to the terminal 22 b of the second operational amplifier 22 , for instance equal to V O1 , which is the voltage value present on the output O 1 of the Hall cell 3 . Since, as has been said, V O1 and V O2 both have the value Vdd/2, in this condition current does not flow through the feedback resistor 26 and the output of the magnetic field sensor 1 is the output 22 c of the second operational amplifier 22 , equal to Vdd/2.
As is known, the presence of a magnetic field B orthogonal to the plane in which the Hall cell 3 lies causes a deviation of the electrical charges that flow between the first terminal 2 and the second terminal 4 . This deviation entails onset of a preferred path for the charge carriers, which will thus flow between the first terminal 2 and the second terminal 4 preferably through the resistors 10 and 12 , or else preferably through the resistors 14 and 16 , according to the orientation of the magnetic field B. There follows an unbalancing between the values of voltage V O1 and V O2 on the third terminal 6 and the fourth terminal 8 , respectively. Since the inverting terminal 22 b of the second operational amplifier 22 is at the voltage V O2 =V O1 =Vdd/2, the second operational amplifier 22 operates for bringing the voltage on the inverting terminal 22 a to the same value V O2 =V O1 =Vdd/2. This results in a current that flows between the output 22 c of the second operational amplifier 22 and the inverting input 22 a , through the feedback resistor 26 , thus generating a non-zero voltage V R _ OUT across the feedback resistor 26 . The value of the current that flows in the feedback branch 24 is such as to compensate the unbalancing between the voltages V O1 and V O2 generated following upon application of the magnetic field B. Consequently, the value of the voltage V R _ OUT read across the feedback resistor 26 indicates a value of the magnetic field B, which may thus be easily measured. Furthermore, the sign of the voltage V R _ OUT across the feedback resistor 26 indicates the orientation of the magnetic field B.
›DETAILED DESCRIPTION OF THE DRAWINGS · 3 of 5
However, as is known, the resistors 10 , 12 , 14 , and 16 are not ideal, and consequently they may show an effective value of the respective resistance R 1 -R 4 that is not uniform. Furthermore, also the first and second operational amplifiers 20 , 22 are not ideal and consequently, for each of them, the difference of potential between the inverting terminal 20 a , 22 a and the respective non-inverting terminal 20 b , 22 b is other than zero. These non-idealities cause a flow of current through the feedback resistor 26 , even in conditions of absence of an external magnetic field B, that is such as to generate a non-zero feedback current through the feedback resistor 26 even in conditions of absence of magnetic field; this voltage is an “offset” of the magnetic field sensor 1 . The offset is an important parameter that affects the performance of the Hall sensor 1 , as is well known in the microelectronic field.
To overcome this drawback, it is advisable to set the voltage on the non-inverting terminal 22 b of the second operational amplifier 22 at a value different from that of the voltage V O1 present on the third terminal 6 of the Hall cell 3 . In particular, it is expedient to choose a voltage value V OFFSET , to be supplied to the non-inverting terminal 22 b , such as to compensate the non-idealities of the resistors 10 , 12 , 14 , 16 and of the operational amplifiers 20 , 22 . For instance, it is possible to choose the voltage value V OFFSET as the value for which, in the absence of magnetic field B that acts upon the Hall cell 3 , no current flows through the feedback resistor 26 , or else to set the voltage value V OFFSET for having on the output 22 c of the second operational amplifier 22 a voltage value, with zero magnetic field B, that is pre-determined and desired (for example chosen according to the input dynamics of the possible circuit cascaded to the Hall sensor, which receives and uses said information).
For instance, according to one embodiment, it is possible to choose the voltage V OFFSET to be supplied to the terminal 22 b of the second operational amplifier 22 of a value such that, in the absence of external magnetic field B, the voltage on the output terminal 22 c has a value equal to V force . Knowing the voltage value of the signal on the output terminal 22 c in the absence of magnetic field B, it is thus possible to choose the voltage value of the signal V OFFSET as the value for which the signal on the output terminal 22 c has the ideal value that it would have in the absence of non-idealities (i.e., V OFFSET is chosen of a value such as to bring the output 22 c of the second operational amplifier 22 to the value V force ). The signal V R _ OUT present across the feedback resistor 26 in this operating condition is chosen as zero magnetic field signal, on the basis of which the variations caused by the presence of the external magnetic field B are measured to evaluate the intensity thereof.
In particular, according to one embodiment, the voltage value V OFFSET is a function of the voltage value V O1 on the third terminal 6 , and is equal to V O1 +V X , where V X is a voltage value chosen for obtaining on the output 22 c of the second operational amplifier 22 , with zero magnetic field B, the ideal voltage value (e.g., Vdd/2). In any case, V O1 is dependent upon the voltage V force , and consequently it follows that, considering the non-idealities of the first operational amplifier 20 , the voltage V OFFSET is a function of the voltage V force .
According to a further embodiment, the non-idealities are compensated by a technique of dynamic offset cancellation, known as “current spinning”. The current-spinning technique is, for example from P. J. A. Munter, “A low-offset spinning-current Hall plate”, Sensors and Actuators, A21-A23, pp. 743-746, 1990, in a per se known manner (document incorporated by reference).
FIG. 3 shows the magnetic field sensor 1 in a configuration such as to implement the current-spinning technique.
The Hall sensor 1 of FIG. 3 comprises, in addition to the elements already illustrated and described with reference to FIG. 2 (in particular, the Hall cell 3 , the first operational amplifier 20 , the second operational amplifier 22 ), and a spinning circuit 32 including a plurality of switches that are designed to implement spinning of the current i BIAS supplied to the terminals of the Hall cell 3 .
The spinning circuit 32 comprises four switch modules 34 , 36 , 38 and 40 . The switch module 34 is electrically coupled to the first terminal 2 of the Hall cell 3 . The switch module 36 is electrically coupled to the second terminal 4 of the Hall cell 3 . The switch module 38 is electrically coupled to the third terminal 6 of the Hall cell 3 . The switch module 38 is electrically coupled to the fourth terminal 8 of the Hall cell 3 .
The switch modules 34 , 36 , 38 and 40 are, from an electrical standpoint, the same as one another. In fact, each switch module 34 , 36 , 38 and 40 includes a respective first switch C 1 and a respective second switch C 2 , which each have a respective conduction terminal electrically coupled to the same terminal 2 , 4 , 6 and 8 of the Hall cell 3 .
During a first operating condition of the spinning circuit 32 , the switches C 1 of each switch module 34 , 36 , 38 and 40 are controlled simultaneously in an ON state (i.e., they are closed), whereas the switches C 2 of each switch module 34 , 36 , 38 and 40 are controlled simultaneously in an OFF state (i.e., they are open). During a second operating condition of the spinning circuit 32 , the switches C 1 of each switch module 34 , 36 , 38 and 40 are controlled simultaneously in an OFF state (i.e., they are open), whereas the switches C 2 of each switch module 34 , 36 , 38 and 40 are controlled simultaneously in an ON state (i.e., they are closed).
The spinning circuit 32 further comprises additional switch modules 42 , 44 , 46 , and 48 , which from a circuit standpoint are the same as one another and each of which includes a respective first switch A and second switch B.
›DETAILED DESCRIPTION OF THE DRAWINGS · 4 of 5
For instance, the switch A of the switch module 42 has the function of coupling/decoupling the current generator 30 to/from the switch C 1 of the switch module 34 (and via the latter to/from the first terminal 2 ). Further, the switch A of the switch module 42 has the function of coupling/decoupling the current generator 30 to/from the switch C 2 of the switch module 38 (and via the latter to/from the third terminal 6 ).
The switch B of the switch module 42 has, instead, the function of coupling/decoupling the current generator 30 to/from the switch C 1 of the switch module 36 (and via the latter to/from the second terminal 4 ). Further, the switch B of the switch module 42 has the function of coupling/decoupling the current generator 30 to/from the switch C 2 of the switch module 40 (and via the latter to/from the fourth terminal 8 ).
Likewise, the switches A and B of the switch module 44 have the function of coupling/decoupling the ground terminal GND alternatively to/from the first input 2 , to/from the second input 4 , to/from the first output 6 , or to/from the second output 8 , via the further control of the respective switches C 1 and C 2 of the switch modules 34 , 36 , 38 and 40 .
In this way, in use, it is possible to make the supply of the biasing current i BIAS to spin between the terminals 2 , 4 , 6 , 8 of the Wheatstone bridge of the Hall cell 3 for providing four different paths for the biasing current I BIAS .
Likewise, the switches A and B of the switch modules 46 and 48 are controlled, in the first and second operating conditions of the spinning circuit 32 , for coupling/decoupling the first and second operational amplifiers 20 , 22 alternatively to/from all the terminals 2 , 4 , 6 , 8 of the Wheatstone bridge of the Hall cell 3 for acquiring four different output voltages.
By way of example, the first operating condition corresponds to a situation in which all the switches C 1 are in the ON state and all the switches C 2 are in the OFF state. In this condition, when also all the switches A are on, we have the situation of FIG. 2 . The biasing current I BIAS flows from the first terminal 2 (input I 1 ) to the second terminal 4 (input I 2 ). The first operational amplifier 20 has its inverting terminal coupled to the third terminal 6 of the Hall cell (output O 1 ), and the second operational amplifier 22 has its inverting terminal coupled to the fourth terminal 8 of the Hall cell (output O 2 ).
Remaining in the first operating condition, by turning off the switches A and turning on the switches B, a situation is obtained whereby the flow of biasing current I BIAS is reversed; i.e., the biasing current I BIAS flows from the second terminal 4 (which thus provides the input I 1 ) to the first terminal 2 (which thus provides the input I 2 ), coupled to the reference terminal (e.g., GND). The first operational amplifier 20 has, in this case, its inverting terminal coupled to the fourth terminal 8 of the Hall cell 3 (which thus provides the output O 1 ), and the second operational amplifier 22 has its inverting terminal coupled to the third terminal 6 of the Hall cell (which thus provides the output O 2 ).
The second operating condition corresponds to a situation in which all the switches C 2 are in the ON state and all the switches C 1 are in the OFF state. In this condition, when also all the switches A are on, the biasing current I BIAS flows from the third terminal 6 of the Hall cell 3 to the fourth terminal 8 of the Hall cell 3 (the latter terminal is coupled to ground GND). The first operational amplifier 20 has its inverting terminal coupled to the second terminal 4 of the Hall cell 3 , and the second operational amplifier 22 has its inverting terminal coupled to the first terminal 2 of the Hall cell 3 .
Remaining in the second operating condition, by turning off the switches A and turning on the switches B, a situation is obtained whereby the flow of biasing current I BIAS is reversed; i.e., the biasing current I BIAS flows from the fourth terminal 8 of the Hall cell 3 to the third terminal 6 of the Hall cell 3 (the third terminal 6 is in fact coupled to ground GND). The first operational amplifier 20 has, in this case, its inverting terminal coupled to the first terminal 2 of the Hall cell 3 , and the second operational amplifier 22 has its inverting terminal coupled to the second terminal 4 of the Hall cell 3 .
In this way, the biasing current has traversed all the possible paths between the terminals 2 , 4 , 6 , 8 of the Hall cell 3 . Likewise, also the voltage supplied to the inverting terminal 22 a of the second operational amplifier 22 has been acquired on each terminal 2 , 4 , 6 , 8 of the Hall cell 3 , for each current path.
For each of the four current paths described previously the voltage signal V R _ OUT is acquired across the feedback resistor 26 . There are thus acquired voltage signals that indicate the magnetic field B that acts upon the Hall cell 3 and that are further affected by an offset generated by the non-idealities of the resistors 10 - 16 and of the operational amplifiers 20 , 22 . The appropriate combination (e.g., sum or temporal average) of the four voltage signals V R _ OUT thus acquired enables a result signal that is without offset to be obtained. In fact, by making the current to spin from one contact to another of the Hall sensor according to the current-spinning technique, we have that the signal due to the magnetic field B and the offset alternatively change sign. By adding said signals together a sum is thus obtained of the useful signals and a difference between the offset components, according to the known operation of the circuits used for current spinning.
The magnetic field sensor 1 of FIG. 2 or FIG. 3 may be obtained in an integrated form in a die comprising an appropriately doped semiconductor or insulating substrate. The operational amplifiers 20 , 22 may coexist on the same substrate. Alternatively, the amplifiers 20 , 22 may be provided on a different substrate and operatively coupled to the Hall cell 3 subsequently.
›DETAILED DESCRIPTION OF THE DRAWINGS · 5 of 5
According to a further aspect of the present invention ( FIG. 4 ), a plurality of Hall cells 3 , 3 ′ (here, two Hall cells 3 , 3 ′) may be coupled together in parallel. The Hall cell 3 ′ is of the same type as the Hall cell 3 .
In this case, a plurality of Hall cells 3 , 3 ′ of the type shown in FIG. 2 have their respective first input terminals 2 , 2 ′ coupled together to a common node 50 , and their respective second input terminals 4 , 4 ′ coupled together to a common node 52 (coupled to the reference terminal, for example at ground voltage GND).
The first output 6 , 6 ′ of each Hall cell 3 , 3 ′ is coupled to the inverting terminal 20 a of the first operational amplifier 20 , whereas the second output 8 of each Hall cell 3 is coupled to the inverting terminal 22 a of the second operational amplifier 22 .
In this way, we have the advantage that the signal V R _ OUT across the resistor 26 is a signal that is a function of the magnetic field B to which all the Hall cells 3 , 3 ′, are subject and has a value higher than in the case of FIG. 2 , where just one Hall cell 3 is present. The current that flows through the feedback resistor 26 is thus the sum of the currents that would be generated in feedback by the second operational amplifier 22 for each Hall cell 3 , 3 ′ considered individually. Consequently, the signal V R _ OUT indicates the sum of the feedback currents and is thus an amplified signal.
The non-inverting terminal 20 b of the first operational amplifier 20 is biased at the voltage V force (e.g., Vdd/2), whereas the non-inverting terminal 22 b of the second operational amplifier 22 may be biased at the voltage V O1 present on the first output 6 or else at the reference voltage V OFFSET chosen as described previously.
To compensate the offsets, it is further possible to couple operatively, to each Hall cell 3 of FIG. 4 , or to the set of the Hall cells 3 in parallel, a spinning circuit 32 of the type described previously.
The present invention finds, for example, application in a system for accurate measurement of the currents that flow in a conductor, e.g., a wire or a conductive structure provided in integrated electronic circuits, either to detect appropriate electrical quantities, correlated to the current flowing in the integrated electronic circuit, or to monitor operation of the electronic circuit itself correctly. The magnetic field sensor 1 may thus be used for measuring the current flowing in said wire, which, for example, forms a coil, which generates, as is known, a magnetic field correlated to the current in the wire itself. The deviation of the charges in the Hall cell 3 , generated by the Hall effect as discussed previously, is detected as potential on the feedback resistor 26 and represents a measurement of the current flowing in the wire.
The advantages according to the present disclosure emerge clearly from the foregoing description.
In particular, the presence of the first operational amplifier 20 fixes the voltage on the terminal 6 of the Wheatstone bridge at the pre-set value V force . In this way, even in the presence of a variable magnetic field that acts upon the Hall cell 3 , we do not have phenomena of charge and discharge of the parasitic capacitance on the terminal 6 . The same reasoning applies as regards the terminal 8 . Also here, the presence of the second operational amplifier 22 coupled to the terminal 8 of the Wheatstone bridge has the function of fixing the voltage on the terminal 8 at the value V O1 or V OFFSET present on the other input of the second operational amplifier 22 . In the presence of the same variable magnetic field acting upon the Hall cell 3 , we do not have phenomena of charge and discharge of the parasitic capacitance on the terminal 8 . The bandwidth of the entire circuit is considerably improved.
The presence of various alternatives for compensating/reducing the offsets due to the non-idealities of the elements that provide the circuit further guarantees freedom of implementation of the preferred offset-compensation procedure, according to the needs and preferences of the designer or to the system in which the present invention is implemented.
Finally, it is clear that modifications and variations may be made to what has been described and illustrated herein, without thereby departing from the scope of the present invention, as defined in the annexed claims.
For instance, it is possible to couple an analog-to-digital converter (not shown) across the feedback resistor 26 , for acquiring the (analog) voltage signal generated following upon the flow of feedback current through the resistor 26 and generating a digital signal.
Claims as granted
10 claimsLog in to read the claims of this application.
Log in to unlockClassifications
3 codes- G01R33/07
- G01R33/00
- G01R33/06
Claim changes
SoonSee which claims were amended, added or cancelled during examination, with every added and removed word marked.
The published claims of this application are not paired with the granted ones in what we hold.
File wrapper
See the full prosecution history — every USPTO and applicant action on this file, in order.
Log in to unlockDocuments
Log in to open the documents of this file: the application as filed, every office action and response, the notice of allowance.
Log in to unlockChain of title
See the full assignment history — every owner this patent has passed through, with recordation dates and reel/frame numbers.
Log in to unlock