Linearization of magnetic sensor output based on continuous correction of higher-order voltage output components

Abstract

To improve the linearity of a magnetoresistive sensor without reducing its sensitivity. The present disclosure relates to a method for correcting an output signal provided by a magnetoresistive sensor in the presence of an external magnetic field, the method comprising: Determining a deviation of the output signal from a linear response due to an amplitude of a higher order component signal of the output voltage; and determining a correction output signal by compensating the output signal of the higher order component signal such that the correction output signal varies in proportion to changes in the external magnetic field within a variation range. The present disclosure further relates to an integrated circuit (IC) configured to perform a method for deriving a common variable used in performing the correction method and a characterization method for a plurality of magnetoresistive sensors.

Description

[Technical Field]

[0001] This disclosure relates to a correction method for correcting an output voltage signal provided by a tunnel magnetoresistive sensor in the presence of an external magnetic field, and an integrated circuit (IC) configured to perform this method. The disclosure further relates to a characterization method for multiple magnetoresistive sensors for deriving common variables used when performing the correction method. [Background technology]

[0002] Linear magnetic sensors have many applications in consumer, industrial, and automotive sectors. Examples include current detection, positioning, proximity detection, and biometric authentication. Sensor technology using magnetic tunnel junctions (MTJs) based on the tunnel magnetoresistance (TMR) effect (hereinafter referred to as TMR sensors) is superior to competing technologies based on the AMR effect, giant magnetoresistance (GMR) effect, and Hall effect due to its high sensitivity, signal-to-noise ratio (SNR), temperature dependence, long-term stability, and overall miniaturization of die size.

[0003] A TMR sensor may comprise one or more magnetoresistive elements, each of which is equipped with a MTJ (Magnetoresistive Transistor Junction). The MTJs are connected in various series and parallel combinations to meet the requirements of a particular application, such as bandwidth, power consumption, and noise. Generally, such a TMR sensor is configured in a Wheatstone bridge arrangement and outputs an output voltage (V) that is approximately proportional to a given external magnetic field. out ) provides. However, the larger the magnetic field, the greater the V from the perfectly linear response. out The deviation becomes larger.

[0004] The linearity of such magnetoresistive sensors can generally be improved by developing new magnetic stacks that allow for a wider operating magnetic field range. However, this improvement in linearity usually comes at the expense of reduced sensor sensitivity.

[0005] External magnetic field (H)V out The typical response of the TMR sensor below can be approximated by the following equation.

number

[0006] Here, a0 is the sensor offset, and a1 and a3 are the coefficients of the linear and cubic components, respectively. Typically, ai >> a3 means that higher-order components (5th, 7th, 9th, ...) can be ignored and are not considered here. The approximation given in Equation 1 is based on measurements of many TMR sensors with different magnetic stacks and has been found to accurately reflect the behavior of the sensors for the purposes of this disclosure.

[0007] Figure 1 shows the linearity error obtained from linear TMR sensors for different magnetic field ranges. out When this is fitted to a linear function, the linearity error increases rapidly as the considered magnetic field range (see dashed line in Figure 1) reaches less than 1% for magnetic fields below 40 mT. out This rapid increase in linearity error due to the presence of additional higher-order components limits the operating magnetic field range of such sensors. Therefore, the ratio of the cubic coefficient to the linear coefficient (a3 / a1) determines the linearity error of the sensor in the fixed magnetic field range or the working magnetic field range, and obtaining a linearity error below a certain value (see Figures 2a and 2b). In Figure 1, the black dots

number

number

[0008] Figure 2a shows a simulation of the linearity error versus the a3 / a1 ratio in a magnetic field range of 100 mT, and Figure 2b shows the maximum magnetic field range when the linearity error versus the a3 / a1 ratio is less than 0.5%.

[0009] Commercially available linear TMR sensors typically operate at a maximum of 40 mT, but there are some applications where high precision (<0.1%) linear response is required (such as precise positioning in surgery or aerospace), or where larger magnetic fields (up to 100 mT) are involved.

[0010] Therefore, a high linear V out Developing an MTJ stack that guarantees response can improve the linearity of the sensor, but at the expense of sensor sensitivity. Lookup table-based solutions or solutions based on the calculation of correction polynomials require an ADC, DAC, memory, and microcontroller, V corr Because it involves a complete digital reconstruction, it consumes more power, is slower, and requires a larger die area. [Overview of the project]

[0011] Two different strategies are possible for developing high-proximity TMR sensors. One is to develop different magnetic stack configurations. The other is to develop a correction strategy to reduce the linearity error of the TMR sensor's output voltage. Each strategy has its advantages and disadvantages, as summarized in Table 1.

[0012] This disclosure discusses a method for correcting the output voltage and improving the linearity of a TMR sensor without reducing sensitivity. Several approaches are proposed to approximately determine and compensate for the higher-order terms of the output voltage, which are the main cause of the nonlinearity of the output voltage response.

[0013] [Table 1]

[0014] The method proposed here enables a significant improvement in linearity error (thus achieving a wider magnetic field range) without compromising sensitivity. Furthermore, the correction method has the potential to be implemented in all linear magnetoresistive sensors, significantly improving the linearity error of existing devices.

[0015] The goal of the correction method described here is to achieve a stable output voltage response that is relatively insensitive to variations between samples, temperature, and operating voltage. This means that such correction can be achieved by applying the same set of variables to all devices on the wafer, avoiding time-consuming individual calibration procedures for each sensor device, which can impact manufacturing costs and reliability performance.

[0016] In particular, this disclosure relates to a correction method for correcting an output signal provided by a magnetoresistive sensor in the presence of an external magnetic field, and this correction method is To determine the deviation of the output signal from the linear response due to the amplitude of the higher-order component signals of the output signal, For magnetic field ranges up to 100 mT, the corrected output signal is determined by compensating the output signal for higher-order component signals such that the corrected output signal has a linearity error of less than 2%, preferably less than 1%, and more preferably less than 0.5%. It is equipped with.

[0017] This disclosure further relates to an integrated circuit (IC) configured to perform a method for deriving common variables used when performing a correction method for multiple magnetoresistive sensors, and a method for characterizing them.

[0018] Exemplary embodiments of the present invention are disclosed in this description and illustrated by the drawings. [Brief explanation of the drawing]

[0019] [Figure 1] Figure 1 shows the uncorrected linearity error and the linearity error after correcting for third-order nonlinearity. [Figure 2] Figures 2a and 2b show simulations of the linearity error versus the ratio of the first-order coefficient to the third-order coefficient (a1 / a3). Figure 2a shows the case for a magnetic field range of 100 mT, and Figure 2b shows the maximum magnetic field range versus a1 / a3 to achieve a linearity error of less than 0.5%. [Figure 3]Figures 3a to 3d show potential implementations of the first correction method, where Figure 3a shows an example of an ASIC circuit for linearity correction, Figure 3b shows a comparison of the magnetic field dependence of the sensor's raw output voltage and corrected output voltage, Figure 3c shows a comparison of linearity errors, and Figure 3d shows an alternative ASIC circuit for linearity correction. [Figure 4] Figure 4 shows the implementation circuit for discontinuous piecewise linear correction. [Figure 5] Figure 5 shows a segmented linear nonlinearity correction method and circuit having three sections according to one embodiment. [Figure 6a] Figure 6a shows the reduction of nonlinearity in a magnetoresistive sensor using a three-part linear correction method. [Figure 6b] Figure 6b shows the reduction of nonlinearity in a magnetoresistive sensor using a three-part linear correction method. [Figure 7] Figure 7 shows the nonlinearity correction for four different magnetoresistive sensors from the same wafer using the same circuit variables. [Figure 8] Figure 8 shows a piecewise linear nonlinearity correction method and a circuit with five sections according to an embodiment. [Figure 9a] Figure 9a shows the reduction of nonlinearity in an actual sensor using the 5-part linear correction method, and shows the nonlinearity of the sensor with and without correction (Figure 9a). Sensor output voltage with and without correction (Figure 9b). [Figure 9b] Figure 9b shows the reduction of nonlinearity in an actual sensor using the 5-part linear correction method, and displays the sensor output voltage with and without correction (Figure 9b). [Figure 10] Figure 10 shows the nonlinearity correction for four different magnetoresistive sensors from the same wafer, using the same circuit variables. [Figure 11] Figure 11 shows the stability of nonlinearity correction in a ratiometric system over a temperature range of 50°C to 150°C and a power supply voltage range of 4.5V to 5.5V. [Figure 12] Figure 12 shows a simplified and preferred embodiment of a piecewise linear nonlinearity correction method and circuit having three sections. [Figure 13a]Figure 13 reports the simplified three-part linear nonlinearity correction of the actual sensor and shows the nonlinearity as a function of the externally given magnetic field (Figure 13a). [Figure 13b] Figure 13 reports the simplified three-part linear nonlinearity correction of the actual sensor and shows the output voltage as a function of a given external magnetic field (Figure 13b). [Figure 14] Figure 14 shows simplified three-part linear nonlinear corrections for four different magnetoresistive sensors on the same wafer. [Figure 15] Figure 15 shows multiple shifts in nonlinearity (NL) correction due to temperature and power supply voltage. [Figure 16] Figures 16a to 16d report the voltage response of a linear magnetoresistive sensor as a function of the magnetic field. Figure 16a shows the linear fit of the output voltage, Figure 16b shows the linear error magnetic field, Figure 16c shows a comparison of the output voltage before and after correction for two different correction variables, and Figure 16d shows the linear error of the corrected output voltage for two different correction variables. [Figure 17] Figure 17 shows the performance of the linearity error based on the linear error correction method of "linear fitting". [Figure 18] Figures 18a to 18c illustrate the verification of such linearity error correction in a linear TMR sensor. Figure 18a shows the raw output voltage as a function of the magnetic field, Figure 18b shows the linearity error of a linear MTJ sensor as a function of the magnetic field, and Figure 18c shows the linearity error considering the output voltage fitted by a cubic polynomial function and the linearity error after output voltage correction using the "linear fitting" correction method. [Figure 19] Figure 19 shows a possible embodiment of "linear fit" linearity error correction using a four-quadrant multiplier, the correction being based on Equations 110 and 107b. [Figure 20] Figure 20 shows another embodiment of the correction of the linearity error for "linear fitting" using a one-quadrant multiplier, the correction being based on Equations 110 and 107b. [Figure 21] Figure 21 shows another embodiment of "linearly fitted" linearity error correction using a one-quadrant multiplier, the correction being based on Equations 110 and 107b. [Figure 22] Figures 22a and 22b show analog IC units based on combinations of logarithmic ratio, logarithmic, and inverse logarithmic operational amplifiers (Figure 22a), and analog IC units as analog multi-purpose units (AMUs). [Figure 23] Figures 23a and 23b show the output voltage and corrected output voltage of an MTJ-based sensor using an AMU (Figure 23a) and a "linear fit" linearity error correction scheme (Figure 23b). [Figure 24] Figure 24 shows an IC according to one embodiment. [Figure 25] Figure 25 reports the linear error obtained from the linear TMR sensor implemented by the IC in Figure 24, after correction of the linearity error, as a function of the IC's input voltage. [Figure 26] Figure 26 shows a fully analog MTJ sensor and ASIC system according to one embodiment. [Figure 27] Figure 27 shows a fully analog MTJ sensor and ASIC system according to another embodiment. [Figure 28] Figures 28a to 28c show the digital implementation of linearity error correction (Figure 28a), simulations of "linear fit" linear correction using 12-bit and 8-bit ADCs for MTJ sensors given a magnetic field of up to 67 mT (Figure 28b), and the linearity error with respect to the number of bits of the ADC (Figure 28c). [Figure 29] Figure 29 shows a flowchart illustrating the characterization and implementation of linearity error correction for a sensor device within a wafer according to an embodiment. [Figure A1] Figure A1 compares the output voltage obtained using equation 103a and an approximate value of equation 103a. [Modes for carrying out the invention]

[0020] The voltage response of a linear TMR sensor can be described by Equation 1, and can be rewritten as follows:

number

[0021] The proposed non-linearity correction method described below depends on the compensation of the high-order components of V out . That is, the corrected output voltage V corr is determined by compensating the output voltage V corr with respect to the high-order component voltage V ho so that the corrected output voltage V out varies proportionally to the variation of the external magnetic field (H) within a larger magnetic field range. This correction may be performed piecewise linearly or continuously.

[0022] This compensation method can be implemented in hardware (analog), software (digital), or a hybrid hardware and software (analog and digital) circuit.

[0023] First Correction Method Piecewise Linear Correction

[0024] The first correction method to be described is the piecewise linear correction method. To explain this approach, the output voltage V out of the sensor is divided into non-overlapping output voltage segments V out,i . This method can be extended to a practical number of output voltage segments. In this explanation, for clarity, the case of three segments is considered first. V out < V1 includes the output voltage segment I (V out,1 ) V out > V2 includes the output voltage segment II (Vout,2 ) V1 ≤ V out ≤ V2, there is an output voltage section III (V out,3 ) Here, when V1 < V2, the transition thresholds V1 and V2 for each section divide the output voltage sections V out,1 and V out,2 and V out,3 into sections.

[0025] Within each output voltage section V out i is approximated by a linear equation.

Equation

number

[0029] In the example in Figure 3a, a pair of comparators operate based on the segmental transition thresholds V1 and V2, determining the output voltage segment V out The parameters ,i are determined, and a corresponding correction signal Vcorr,i is output accordingly. Correction function (Ai+Bi*V out This can be easily implemented in common analog circuits such as operational amplifiers and passive components. The example in Figure 3d uses a comparator to determine the output voltage segment V out,i This is an alternative implementation that selects pairs of Ai and Bi coefficients based on the following. This concept is illustrated in a three-part scenario, and naturally, more output voltage parts V out,i This can be expanded to more accurately compensate for the nonlinearity of the sensor. Note that the circuit shown in Figure 3a can be equipped with only one comparator 10 (for example, when only half of the sensor output is used, such as in unipolar applications).

[0030] A particularly useful implementation of equation 102a is shown in the circuit diagram in Figure 4, and the following should be considered. V out If the value is small, no correction is necessary. A3=0 and B3=1, therefore V corr,3 =V out,3 Therefore, the circuit in Figure 4 may have only one comparator 10.

[0031] Corrected output voltage V corr Since discontinuities are highly undesirable in the application, discontinuities should not be introduced during segmental transitions. V1 and V2 are segmental transition voltages, which can be achieved by maintaining the relationship between Bi and Ai as follows:

number

[0032] A preferred embodiment of the piecewise linear correction method shown in Figure 4 comprises conventional circuit elements such as operational amplifiers, transistors, and resistors, as shown in the circuit of Figure 5. In this preferred embodiment, the functions of the comparator and voltage source in Figure 4 are combined in a voltage-to-current (V-to-l) converter composed of an operational amplifier, a MOS transistor, and a resistor. The addition operation is performed in the current domain by a current mirror to which the output current is supplied to R0, and the corrected output voltage V corr Generates. The segmental transition threshold voltages V1 and V2, as well as R1 and R2, are preferably implemented as programmable variables that can be changed based on the sensor characteristics to optimize nonlinearity correction. Output signal segment V out,i If the value is greater than the segmentation transition threshold Vi, the output signal segment V out,i Outputs, output signal segment V out,i The threshold for the transition between segments is V i Output signal segment V when it is greater out,i Corrected output signal segment V added to corr,i It is configured to output the following.

[0033] Following Figure 5, from one perspective, the first voltage-current conversion circuit 15a includes a first resistor R1 and a voltage signal V out,i and threshold signal V i It can be configured to generate a first current as a function of the difference between and . The second voltage-current conversion circuit 15b includes a second resistor R2 and can be configured to generate a second current i2 as a function of the first current i1. When a second current i2 is supplied to the correction resistor R0, the correction output signal segment V corr,iThis circuit generates the output signal segment V. out,i The threshold for the transition between segments is V i Output is generated when the latter is smaller, and the latter is the segmentation transition threshold V i Output signal segment V when it is greater out,i Corrected output signal segment V added to corr,i It is configured to output the first current i1, which is the output signal segment V. out,i It can be generated as a linear function of the difference between and the segmental transition threshold Vi. The circuit in Figure 5 may have only one voltage-to-current conversion circuit (such as 15b) for unipolar application.

[0034] Figures 6a and 6b show the reduction in nonlinearity of an actual magnetoresistive sensor using a three-part piecewise linear correction method within a magnetic field range of -45mT to +45mT. Figure 6a shows the nonlinearity of the sensor with and without correction. Figure 6b shows the sensor output voltage with and without correction. As can be seen from the plots, the three-part piecewise linear correction method (from -1.3% to -0.25% of the full scale (measurement range)) reduces the nonlinearity by approximately five times. In this case, the following circuit variables

number

[0035] Typically, magnetoresistive sensors from the same wafer exhibit similar nonlinear characteristics. Therefore, the correction circuit variables are determined once per wafer and can be applied to all sensor dies on the same wafer. Figure 7 shows the nonlinearity correction for four different sensors from the same wafer using the same circuit variables. As observed in the plot, the nonlinearity cancellation is effective for all presented sensors. In Figure 7, the three-part nonlinearity correction for four magnetoresistive sensors from the same wafer is the same set of circuit variables as in Figure 6.

number

[0036] The piecewise linear nonlinearity correction method shown in Figure 4 can be easily extended to more output voltage segments to achieve a higher level of nonlinearity correction. Figure 8 shows a preferred embodiment with five segments.

[0037] Figures 9a and 9b show the reduction in nonlinearity of an actual magnetoresistive sensor using a 5-segment piecewise linear correction method within a magnetic field range of (minus) -45mT to +45mT. Figure 9a shows the nonlinearity of the sensor with and without correction. Figure 9b shows the sensor output voltage with and without correction. As can be observed in the plots, a reduction of approximately 9 times in nonlinearity is achieved with the 5-segment piecewise linear correction method (from approximately 1.3% to approximately 0.14% of the measurement range). In this case, the following circuit variables were used.

number

[0038] Please note that the sensors examined in Figures 9 and 8 are the same as the sensors examined in Figure 6.

[0039] Figure 10 shows the nonlinearity correction for four different magnetoresistive sensors from the same wafer using the same circuit variables. As observed in the plot, the nonlinearity cancellation is effective for all the sensors presented. In Figure 10, the nonlinearity correction for the five divisions uses the same set of circuit variables as in Figures 9a and 9b.

number

[0040] Embodiments shown in Figures 5 and 8 provide stable nonlinearity correction over temperature and power supply voltage ranges (assuming sensor-specific nonlinearity characteristics do not change over temperature and voltage ranges). Figure 11 shows the stability of nonlinearity correction (considering 5 segments) over temperature ranges from (minus) -50°C to 150°C and power supply voltage ranges from 4.5V to 5.5V in a ratiometric system. Note that in a ratiometric system, segment threshold voltages V1, V2, V3, and V4 also need to be varied ratiometrically with the power supply voltage, which can be easily implemented using a voltage divider. In a non-ratiometric system, segment threshold voltages V1, V2, V3, and V4 need to be constant voltage levels that do not depend on temperature and can be generated by a voltage reference that is not affected by temperature.

[0041] In Figure 11, the following variables were used for the corrected stability of the ratiometric system's nonlinearity across five temperature ranges from -50°C to 150°C and power supply voltage ranges from 4.5V to 5.5V.

number

[0042] In the preferred embodiments shown in Figures 5 and 8, the voltage-current converter should preferably be designed to have a higher slew rate and wider bandwidth than the main signal chain, with a phase margin large enough to avoid overshoot, in order to enable continuous, real-time nonlinearity correction. However, the requirements for gain and input offset are not necessarily stringent, so the design can be done relatively easily.

[0043] Another simpler and preferred embodiment of the piecewise linear correction method shown in Figure 4 comprises traditional circuit elements such as transistors and resistors, as shown in Figure 12. Figure 12 shows a simplified embodiment of the piecewise linear nonlinearity correction method using three sections. In this preferred embodiment, the functions of the comparator and voltage source in Figure 4 are combined in a simplified voltage-to-current (V-to-I) converter consisting of MOS transistors and resistors (each voltage-to-current converter circuit 15a, 15b may also include a MOS transistor). The addition operation is performed in the current domain by a current mirror to which the output current is supplied to R0, and the corrected output voltage V corr Generates.

[0044] A preferred embodiment shown in Figure 12 uses resistors and transistors arranged in a current mirror configuration to obtain bias voltages V1 and V2 and PMOS / NMOS transistor threshold voltage V TP and V TN V is determined by each of these factors. out Within the range, V out It generates a current proportional to the value. Figure 12 shows a simple current mirror based on a MOS transistor, but the same function can be achieved using bipolar junction transistors (BJTs) with different current mirror configurations.

[0045] For simplicity, the equations listed in Figure 12 are those in which i1 and i2 are the exact values ​​of (V1+VTP) and (V2+VTN), respectively. out Although it suggests that current begins to flow at a certain level, the turn-on of the MOS transistor is inherently gradual. This operation of the MOS transistor has the advantage of being able to smooth the transitions between output voltage segments. On the other hand, since the segment threshold voltage depends on the threshold voltage of the MOS transistor, the correction depends on variations in the processing step, temperature, and power supply. Furthermore, true ratiometric correction cannot be established in this simplified embodiment. Nevertheless, this simple circuit can significantly improve linearity. The three-segment configuration circuit shown in Figure 12 can, of course, be extended to more output voltage segments, allowing for a higher level of nonlinearity correction. Figures 13a and 13b show a simplified three-part piecewise linear correction applied to an actual magnetoresistive sensor using the following circuit variables.

number

[0046] Similar to the previous embodiment, the set of correction circuit variables is the same as in Figures 13a and 13b.

number

[0047] Figure 15 shows the shift in nonlinearity correction due to temperature (due to changes in MOS transistor characteristics) and power supply voltage (due to the lack of true ratiometry). Note that the corrected maximum nonlinearity, optimized to maintain less than 0.2% (<) of the measurement range at 27°C and 5V, nearly doubles in the temperature range of -50°C to 150°C and the power supply voltage range of 4.5V to 5.5V. This is in contrast to the previous embodiment, which showed little change over the same temperature and voltage range. In Figure 15, the simple nonlinearity-corrected stability of three segments over a temperature range of (minus) -50°C to 150°C and a supply voltage range of 4.5V to 5.5V in a ratiometric system was calculated using the following set of variables.

number

[0048] Second Correction Method: Linear Fit Correction

[0049] Another possible method is, output signal V out kara-V ho Sufficiently close to (in other words, higher-order component signal V) ho(An additional voltage signal V corresponding to the negative value of) sub It depends on the decision of V. out V sub By adding this, a highly linear corrected output voltage V corr This can be derived. In other words, the output signal V out Additional signal V sub (Output signal V out (Derived from) is added to obtain the higher-order component signal V ho Output signal V out By compensating for this, the corrected output signal V corr This is what is required. For example, V out Assuming that this can be expressed in equation 1, and that a1>>a3 and a5~0 (which is usually the case),

number

[0050] Therefore, V sub is a3·H 3 It is necessary to get as close to it as possible.

number

[0051] To achieve this, the most accurate "estimation" possible of the measured magnetic field H is essential, and it should be noted that solving the cubic equation 10³a to determine H would negatively impact the sensor's time response and power consumption. The idea behind this correction method is to use an approximate solution of the measured field H to V sub a3·H 3 The goal is to significantly reduce linearity errors while minimizing the impact on power consumption and the sensor's time response, as this brings the value close enough to the target value. Approximate for ax that is sufficiently less than 1

number

number

[0052] Note that in this case, the sensor offset term a0 is omitted for clarity. Thus, Equation 104 can be regarded as an approximate description of the magnetic field dependence of the magnetoresistive sensor's V out It should be noted that this approximation means that a much simpler analytical solution can be found than simply deriving the solution of H from Equation 103a. As a result, V out derived from V sub can be determined, and the linearity error can be significantly reduced. The solution of Equation 104 can be approximated as follows (see the attached matter for a complete analysis). [Number] With the following [Number]

[0053] This means that V sub can be described as follows. [Number]

[0054] And the corrected output voltage V corr can be described as follows. [Number]

[0055] This correction method can be somewhat generalized considering the following. [Number] With the following [Number]

[0056] Note that FIGS. 16a to 16d show the performance of this correction method.

[0057] FIG. 16a shows the voltage response as a function of the magnetic field of a linear magnetoresistive sensor having a linear coefficient [Number] and a cubic coefficient [Number] . The gray line shows the linear fit (LinFit) of V out . FIG. 16b shows the linearity error [Number] as a function of the magnetic field. The error induced when considering V out as a perfect linear function reaches 5%. Note that even for small magnetic fields (less than 20 mT), V out shows a linearity error greater than (>) 1% (see FIG. 16b). FIGS. 16c and 16d show the performance of the correction method based on Equations 105 and 107. FIG. 16c shows the comparison of V out (black curve) and the corrected V out (V corr ) for two different values of the variable k (dark gray curve and light gray curve). With this approach, the linearity error can be reduced to a value of < 0.5% (therefore, a reduction by x10). Note that the sensitivity of both V corr signals is approximately 1.5 mV / V / mT, which is the same as the linear coefficient (a-,) of V out . These results support the possibility of such a correction method where no reduction in sensitivity is obtained. FIG. 16d shows the linearity error of both V corr (for two different values of the variable k).

[0058] V according to Equations 105 and 107 subDetermining this may require a large amount of computational power. To overcome this problem, a lower-order solution in equation 105a may also be considered.

number

[0059] Nevertheless, the smaller the degree of the solution, the greater the discrepancy between Ho and the measurement field H, and the larger the linearity error. The optimal compromise between low computational requirements and high linearity error correction can be obtained by considering the following:

number

number

[0060] In fact, if we consider equations 110 and 107b as a correction scheme, then for a magnetic field of up to 94 mT, V corr A linearity error of less than 0.5% can be obtained (see Figure 17). This approach to correcting the linearity error is called "linear fitting" linearity correction.

[0061] one From this perspective, the corrected signal V corr To derive the output signal V out The additional signal V is added to it. sub The output signal V out (V out 3 It is proportional to the cube of ().

[0062] one From this perspective, the additional signal V sub The output voltage signal V out Further including additional terms proportional to components of a higher order than the third-order component, we have:

number

[0063] another In terms of, the additional signal V sub can be further defined as follows. [Number] In the above formula, 0.5 < C < 4, where c1 is the linear coefficient determined by the linear fitting of the output signal (V out ), and a3 is the third-order coefficient of the output voltage.

[0064] another In terms of, the additional signal V sub can be further defined as follows. [Number] Here [Number] is involved where a1 and a3 are the linear coefficient and the third-order coefficient of the output signal V out respectively.

[0065] another In terms of, the additional signal V sub can be further defined as follows. [Number] Here [Number] is involved, where a1 and a3 are the linear coefficient and the third-order coefficient of the output signal V out respectively.

[0066] Figure 17 shows the linear coefficient [Number] third-order coefficient [Number] The performance of the linearity error correction method (considering equations 110 and 107b) for the linear MTJ sensor is shown for the cases k=a3 (dark gray curve), k=(3 / 2)a3 (light gray curve), and k=(5 / 3)a3 (black curve).

[0067] Figures 18a to 18c illustrate the verification of such "linear fit" linearity error correction approaches on different linear TMR sensors. Figure 18 shows V out And its linearity error is shown as a function of a magnetic field up to 67 mT. In this case,

number

[0068] This TMR sensor achieves an initial linearity error of 0.7% for magnetic fields up to 67 mT (see Figure 18b). The linear coefficients derived from such linear fitting are:

number

[0069] Furthermore, this "linear fit" linearity correction is highly robust to typical variable variations between devices. Table 2 summarizes the results for eight magnetoresistive sensors given magnetic fields up to 47 mT. The initial linearity error was approximately 1.35% for all of them, and after the "linear fit" correction, the linearity error decreased to ~0.15%. This improvement in linearity error (approximately 9 times) is achieved by using the same c1 and correction coefficient, despite the initial variance of the c1 and a3 variables (approximately 10%) between devices.

[0070] [Table 2]

[0071] Table 2 shows the results of correcting the linearity error of "linearly fitted" linear TMR sensors when a magnetic field of up to 47 mT is applied. The coefficients c0 and c1 are V out The coefficient obtained by linear fitting, i.e., V out = c0 + c1H, and the coefficients a0, a1 and a3 are

number

number

[0072] Figure 19 shows an embodiment of a "linearly fitted" linearity correction implementation. This IC comprises two cascaded voltage multipliers 12. The multiplier 12 is an analog IC unit composed of a combination of multiple logarithmic and inverse logarithmic operational amplifiers, and its signal output V MULT Since it is the product of two input signals V1 and V2, V MULT=V1V2. By cascading and combining the two multipliers 12, V sub The main component of the signal is (approximately) ~V out 3 This is required. In this embodiment, the multiplier 12 can operate with any possible polarity of V1 and V2 (a four-quadrant multiplier). This circuit also has a gain of G = k / c1 3 It is equipped with an operational amplifier 13.

number

number

number

[0073] The multiplier can only operate with one specific polarity of V1 and V2 (a one-quadrant multiplier), V out If linearity correction is required for both polarities, an alternative embodiment is shown in Figure 20. Without loss of generality, this particular embodiment shows a multiplier 12 that operates only for V less than zero. In this embodiment, the IC has two multipliers 12 and a gain G = k / c1 3 The system comprises an operational amplifier 13, two inverting amplifiers 14, a comparator 10, and at least one of a multiplexer 11 (MUX) and a demultiplexer 11 (DMUX). The comparator, inverting amplifier, and at least one of the MUX and DMUX functions are performed by the voltage multiplier 12, which controls the output voltage V out It operates in both polarities, Vout V for either polarity sub The goal is to ensure that the decision is made reliably. Output signal V out The comparator 10, which receives the input, activates both MUX11s, and depending on the output value of the comparator, the MUX selects one of the two input signals. Therefore, the first multiplexer may be configured such that its output signal is always positive, enabling the operation of at least two cascaded multipliers to control the signal ~V out 3 The signal is calculated. After this signal is amplified by op-amp 13, the signal

number

number

[0074] Adding a multiplier to such a cascade structure results in V out Note that this can correct for other higher-order contributions (5th, 7th, ...).

[0075] V for a 1-quadrant multiplier out Another embodiment of the linearity correction in both polarities is shown in Figure 21. Here, the output voltage signal V out The voltage signal offset V0 is added to the voltage signal offset V0 and the output voltage signal V out Input voltage V corresponding to the sum of in This is input to the two cascade voltage multipliers 12). The idea behind this case is V subBefore determining the output voltage V out A voltage offset (V0) is added, so the input voltage is affected for all magnetic fields.

number

number

[0076] This configuration allows us to remove MUX and DMUX in addition to the comparator considered in the previous embodiment in Figure 20. If we want to involve the same number of multipliers as in Figure 20, according to Equation 111, V0 and V0 2 and V0 3 These need to be three different reference voltages. However, a special case in this embodiment is as shown in Figure 21,

number

[0077] Furthermore, in all the embodiments described above, by adding an additional multiplier to such a cascade structure, V out This allows for correction of other higher-order contributions (5th, 7th, ...).

[0078] In all previous examples, the cascade multiplier is V sub ~V out 3 It was used to obtain [the desired result]. However, other analog IC units can also be considered for this purpose. As depicted in Figure 22a, several analog IC units based on combinations of logarithmic, logarithmic, and inverse logarithmic operational amplifiers can perform the following operations.

Number

[0079] Considering various types of operations (multiplication, division, power, and root) that may be executable, this analog IC unit is defined as an analog multi-purpose unit (or AMU) and is depicted in Figure 22b. The input signals

Number

Number

[0080] Therefore, by adding the AMU to the output signal of the sensor V out , a linearized corrected output signal out

Number

[0081] Figure 23a shows the output voltage V out and the corrected output voltage V corr of the MTJ-based sensor when exposed to a magnetic field from 0 mT to 140 mT obtained by the AMU considering the embodiment of Figure 24. In this specific case

Number

Number

number

number

[0082] Figure 24 shows an IC according to an embodiment that includes a full-bridge magnetoresistive sensor 20 equipped with four magnetoresistive effect elements 2, a differential amplifier 13a, an AMU 14, and a non-inverting summing amplifier 13b.

[0083] Furthermore, Figure 25 reports the linearity error obtained from the linear TMR sensor and the IC described in Figure 24 as a function of the input voltage Vin, 3. Figure 25 shows the linearity error between 1.7V and 2.4V. in,3 This demonstrates that even with fine-tuning, a linearity error of less than 0.45% can be achieved.

number

[0084] All of these results, This linearization correction concept applies to the fully analog MTJ sensor and ASCI system (output voltage V, which depends on the external magnetic field). out An MTJ-based magnetic sensor (Equation 103) that shows, Output voltage V out,AMU ga V out 3 An AMU configured to be proportional to (described in Equation 113), Output voltage

number

[0085] Here, when using the first quadrant AMU, the embodiment of FIG. 24 functions only for one polarity of V out and thus it should be noted that it functions only for one direction of the magnetic field. Next, several other embodiments (FIGS. 26 and 27) may be considered to implement this idea of linearization correction for the positive magnetic field and magnetic field amplitude. For this purpose, two options (similar to those described above in FIGS. 20 and 21) can be considered.

[0086] For example, in one embodiment, a full analog MTJ sensor + ASIC system (see FIG. 26) comprises the following. An MTJ-based magnetic field sensor, A comparator (for determining the polarity of the output voltage V out ), Two inverters (one to invert the polarity of V out and the other to invert the polarity of V out,AMU ), V out,AMU ~V out 3 An AMU that enables the calculation of V sub An appropriate V out and V AMU for selecting signals, at least some of either a MUX or a DMUX, An output voltage is

Number

[0087] In another embodiment, a full analog MTJ sensor + ASIC system (FIG. 27) comprises the following. A magnetoresistive sensor, For all magnetic fields

Number

[0088] It should be noted that in all of the above embodiments, higher-order correction terms (5th, 7th, ...) can also be implemented by adding additional AMUs for n = 5, 7...

[0089] From one perspective, the additional signal V sub The output signal V out It may also have additional terms proportional to higher-order components than the third-order component, as follows:

number

[0090] Table 3 summarizes that while different approaches can be considered depending on the initial a3 / a1 ratio, only the first approach ("linear fitting") is relevant for most analog implementations, as other approaches (such as "2D fitting" or "3D fitting") may imply more complex analog IC systems. In particular, Table 3 shows V with a linearity error of less than 0.5% in the magnetic field range up to 100mT. corrWe report the conditions for the a1 and a3 coefficients to obtain the desired result.

[0091] [Table 3]

[0092] Nevertheless, V sub When considering digital analysis, "2D fitting" or "3D fitting" correction methods can also be considered. Figures 28a to 28c show V sub The signal is basically determined digitally, and this signal is converted to raw V by a DAC. out Because it can be subtracted from, pure analog V corr This shows an approach to obtaining a signal. In this case, the main features of the correction method are (as shown in Figure 28a) 1) Analog-to-digital converter (ADC), 2) V sub A digital system (DS) that determines, 3) Digital-to-analog converter (DAC) It should be noted that such a DS can consist of a microcontroller (MCU), a lookup table (LUT), or other combinations of a microprocessor, memory unit, and MCU.

[0093] Figure 28a shows a diagram of one embodiment for the digital implementation of linear error correction. out The signal is converted to a digital signal by an ADC. sub The decision is made by DS. V sub Once it is determined digitally, it is converted to an analog signal, V out Added to V corr The following results are obtained. Figure 28b shows a simulation of "linearly fitted" linearity correction using 12-bit and 8-bit ADCs for a magnetoresistive sensor given a magnetic field of up to 67 mT. Figure 28c shows the linearity error with respect to the number of bits of the ADC.

[0094] In Figure 28a, the ADC and DAC are V sub Used only for the calculation of V outSubtraction from is performed by analogy. Therefore, a simpler design with fewer bits can operate faster with reasonable power consumption. Furthermore, this approach has the advantage of being able to be combined with either of the proposed analog linearization methods, especially when a large magnetic field is given, and all that is needed is, again, a simpler design and fewer bits. corr Unlike previously developed digital linearization approaches that rely entirely on digital reconstruction, this proposed method uses an additional signal V sub Digital decision and V out and V sub Please note that this is based on analog correction (after digital-to-analog conversion) through the addition of [specific component / feature].

[0095] Finally, for such a correction method to be implemented at the production level, it is necessary not only to demonstrate its robustness against variable variability between devices (as shown in Table 2), but also to derive such common variables c1 without complete characterization of each individual device on the wafer. Once these variables are determined, a common ASIC system can be implemented to perform linearization correction for all devices on the same wafer.

[0096] In one embodiment, the medium is a non-temporary computer-readable medium that stores a program that causes a computer to perform the method described above.

[0097] In one embodiment, a characterization method for deriving common variables for multiple TMR sensors is disclosed, wherein the common variables are used when performing the correction method described above.

[0098] From one perspective, the characteristic evaluation method is Multiple magnetoresistive sensors are provided, and the output signal V from each magnetoresistive sensor out Measuring and Measured output signal V out of

number

number

[0099] Output signal V out The measurement can be performed when the magnetoresistive sensor is irradiated with an external magnetic field H corresponding to the maximum operating magnetic field range H2 of the magnetoresistive sensor.

[0100] Multiple magnetoresistive sensors may comprise a subset of magnetoresistive sensors configured within a wafer. For example, the subset of magnetoresistive sensors may comprise between 10 and N (numbers), where N is the total number of magnetoresistive sensors on the wafer.

[0101] From one perspective, the output signal V out Measurement can be performed when the magnetoresistive sensor is exposed to an external magnetic field H corresponding to at least five different magnetic field magnitudes. The external magnetic field H corresponds to a high amplitude of the magnetoresistive sensor's maximum operating magnetic field range H2 and the output signal V out This can be between the low-amplitude magnetic field range H1, which follows a linear dependence with respect to the magnetic field H.

number

[0102] Therefore, the offset a0 and the linear coefficient a1 are V in the low magnetic field range. out It is obtained by linear fitting. The cubic coefficient can be derived by the following equation.

number

[0103] Figure 29 is a flowchart showing a characteristic evaluation method for obtaining common variables by evaluating only a certain number of sensor devices N (10 < N) of a wafer with only 5 magnetic field points or 5 magnetic field devices. H2 is usually the maximum operating magnetic field range of the sensor, and H1 is a small value of the magnetic field (usually 1 to 6 mT).

[0104] Here, the output signal V described above out , the higher-order component signal V ho , the corrected output signal V corr , the output signal section V out,i , the corrected output signal section V corr,i , the additional signal V sub , the signal offset V0, the threshold signal V i , the input signal V in may be in the form of the voltage or current described above.

[0105] Appendix A

[0106] The solution of Equation 104 can be described as follows. [Number] With the following [Number]

[0107] Furthermore, the maximum magnetic field range in which V out can be approximated to Equation 103a is delimited by the magnetic field at which V out is a local maximum or minimum value (see Figure A1). This magnetic field Hc is obtained by minimizing Equation 103a and is as follows. [Number]

[0108] This is the case for magnetic fields from -Hc to Hc.

number

[0109] Therefore, for the magnetic field range of interest, D < 1, equation A01 can be approximated as follows:

number

[0110] H+ is in a magnetic field

number

[0111] Advantages of the disclosed technology

[0112] The correction method presented here expands the operating magnetic field range of a magnetoresistive sensor by improving its linearity at high magnetic fields, or enables operation within the same magnetic field range with higher linearity without reducing sensitivity.

[0113] Furthermore, the proposed correction method is suitable for real-time correction of nonlinearity using analog means, and therefore enables high-bandwidth operation.

[0114] Analog nonlinearity correction (the first correction method according to the embodiments shown in Figures 3, 5, 8, and 12 does not require a microcontroller, ADC, or DAC, is stable with respect to temperature and power supply voltage, is applicable to the entire wafer, and enables real-time continuous correction that requires a small reach of a magnetoresistive sensor.

[0115] The framework for nonlinearity correction based on Equation 110 and Table 2 (for the second correction method, please refer to the analog implementation embodiments shown in Figures 19, 20, 21, 24, 26, and 27) is as follows: Real-time continuous correction without the need for a microcontroller, ADC, or DAC, Robustness to variations in variables between devices, V out The possibility of digitally implementing this approach for calculating higher-order components (see Figure 29). This makes it possible. The nonlinearity correction method allows for the use of a wafer-level rapid determination method for the two main variables of the "linear fit" correction method (flowchart in Figure 30).

[0116] The technologies disclosed herein are Without needing to develop a new MTJ stack, it is possible to improve the performance (linearity error or magnetic field range) of current linear magnetic sensors, Development of a new linear magnetic sensor product based on a linearity error correction method. This makes it possible.

[0117] Output signal V provided by a magnetoresistive sensor in the presence of an external magnetic field H out The correction method described herein for correcting the magnetic field range up to 100 mT provides a corrected output signal V with a linearity error of less than 2%, preferably less than 1%, and more preferably less than 0.5. corr This allows us to obtain the following. Here, the linearity error is defined as the difference between the output voltage signal, measured as a function of the external magnetic field, and the ideally linear relationship between the output voltage signal and the external magnetic field. [Explanation of Symbols]

[0118] 10 Comparator 11 Multiplexers, Demultiplexers 12 Multipliers, Voltage Multipliers 13 Operational amplifiers, voltage amplifiers 13a First voltage amplifier, differential amplifier 13b Second voltage amplifier, non-inverting summing amplifier 14, 14a, 14b Analog Multipurpose Unit (AMU) 15a First Voltage-Current Conversion Circuit 15b Second Voltage-Current Conversion Circuit 16 transistors 2 magnetoresistance element 20 Magnetoresistive Sensor a0 offset coefficient a11th degree coefficient a33rd order component c0 Approximation Offset Coefficient c1 is the approximation linear coefficient. H external magnetic field H2 Maximum Operating Magnetic Field Range i1 1st current i2 2nd current R0 correction resistor R1 1st resistor R2 second resistor V corr Corrected output voltage V ho Higher-order component signals V in , V in,i Input signal V0 signal offset V out Output signal V out,i Output signal segment V i Transition threshold signal V sub Additional signals

Claims

1. A correction method for correcting the output signal provided by a magnetoresistive sensor in the presence of an external magnetic field, To determine the deviation of the output signal from the linear response due to the amplitude of the higher-order component signals of the output signal, The corrected output signal is determined such that the corrected output signal has a linear error smaller than the linear error of the output signal by compensating the output signal for the higher-order component signal. The system includes determining an additional signal that is close to or equal to the negative value of the aforementioned higher-order component signal. Determining the corrected output signal includes compensating the output signal for the higher-order component signal by adding the additional signal to the output signal. Correction method.

2. The output signal [Math 1] Described by, where H is the external magnetic field, Vho is the higher-order component signal, a0 is the offset coefficient, and a1 is the first-order coefficient, The correction method according to claim 1, wherein the higher-order component signal is described by at least a third-order coefficient a3.

3. The process involves dividing the output signal into a plurality of non-overlapping output signal segments, wherein each output signal segment is distinguished by a segmentation transition threshold, and the process includes the division described above. The correction method according to claim 1, wherein each output signal segment is approximated by a linear equation that obtains a corresponding corrected output signal segment.

4. The correction output signal within each output signal segment is [Math 2] The correction method according to claim 3, wherein Vcorr,i is the correction output signal segment, Vout,i is the output signal segment, and Ai and Bi are segmentation coefficients.

5. Each output signal segment is such that i is an index indicating the i-th segment, d0i is the offset coefficient, and d1i is the linear coefficient. [Math 3] It is approximated by, and here, [Math 4] The correction method according to claim 4.

6. The correction method according to claim 1, wherein the additional signal is proportional to Vout 3.

7. The aforementioned additional signal is Vsub is the additional signal, and a2j+1 is the coefficient that determines the proportionality factor of each 2j+1th order component of the output signal Vout. [Math 5] The correction method according to claim 1, further comprising additional elements proportional to a plurality of higher-order components than the third-order component of the output signal Vout, such as the above.

8. The aforementioned additional signal [Math 6] Defined by, where H0 is [Number 7] Defined by [Number 8] Here, a1 is the linear coefficient of the output signal, a3 is the cubic coefficient of the output signal, and C is a constant, thereby the corrected output signal [Number 9] The correction method according to claim 1, as defined by [the relevant definition].

9. H0 [Number 10] Defined by, here [Math 11] The correction method according to claim 8, wherein a1 is a linear coefficient of the output signal and a3 is a cubic coefficient of the output signal, respectively.

10. The aforementioned additional signal [Math 12] Defined by, here [Number 13] accompanied by The output signal of the correction is [Number 14] As defined by, c1 is a linear coefficient determined by the linear fitting of the output signal with respect to a given magnetic field H, The correction method according to claim 1, wherein a3 is the cubic coefficient of the output signal.

11. The correction method according to claim 7, wherein when determining an additional signal, a signal offset is added to the output signal.

12. A non-transient computer-readable medium storing a program that causes a computer to execute a method for correcting an output signal provided by a magnetoresistive sensor in the presence of an external magnetic field, wherein the correction method is Determining the deviation of the output signal from the linear response due to the amplitude of the higher-order component signals of the output voltage, The corrected output signal is determined by compensating the output signal of the higher-order component signal so that the linearity error of the corrected output signal is smaller than the linearity error of the output signal. The system includes determining an additional signal that is close to or equal to the negative value of the aforementioned higher-order component signal. A non-transient, computer-readable medium storing a program that causes a computer to execute a correction method, wherein determining the corrected output signal comprises compensating the output signal for the higher-order component signals by adding the additional signal to the output signal.

13. An integrated circuit (IC) configured to perform a method for correcting an output signal provided by a magnetoresistive sensor in the presence of an external magnetic field, wherein the correction method is: Determining the deviation of the output signal from the linear response due to the amplitude of the higher-order component signals of the output voltage, The corrected output signal is determined by compensating the output signal of the higher-order component signal so that the linearity error of the corrected output signal is smaller than the linearity error of the output signal. The system includes determining an additional signal that is close to or equal to the negative value of the aforementioned higher-order component signal. An integrated circuit (IC) is configured to perform the correction method, wherein determining the corrected output signal comprises compensating the output signal for the higher-order component signals by adding the additional signal to the output signal.

14. The process involves dividing the output signal into a plurality of non-overlapping output signal segments, wherein each output signal segment is distinguished by a segmentation transition threshold, and the process includes the division described above. Each output signal segment is approximated by a linear equation that yields the corresponding corrected output signal segment. The IC according to claim 13, wherein the output signal segment, the corrected output signal segment, and the segment transition threshold are voltages, the IC comprises at least one comparator, and the output signal and one of the segment transition thresholds are input to the comparator (10).

15. The IC according to claim 14, comprising a multiplexer configured to select one of a plurality of correction output signal segments based on the output of at least one of the comparators.

16. The corrected output signal within the output signal segment is such that Vcorr,i is the corrected output signal segment, Vout,i is the output signal segment, and Ai and Bi are segmentation coefficients. [Number 15] The IC according to claim 14, wherein the comparator is configured to select the classification coefficients Ai and Bi, determined by the comparator.

17. At least one of the comparators' outputs is connected to a correction voltage generator that generates a correction voltage, When the sensor output voltage is less than the segmental transition threshold, the IC outputs the sensor output voltage. The IC according to claim 14, configured to output the sum of the output voltage of the sensor and the correction voltage when the output voltage of the sensor is greater than the classification transition threshold.

18. The first voltage-current conversion circuit is configured to generate a first current, and here The first current is a function of the difference between the output voltage signal of the sensor and the threshold signal when the output voltage signal of the sensor is greater than the threshold signal. The first current is zero when the output voltage signal of the sensor is smaller than the threshold signal. The IC according to claim 13, wherein a correction resistor is configured between the sensor output voltage signal and the correction output signal, which generates a correction output signal when the first current is supplied to the correction resistor.

19. The IC according to claim 18, wherein the first current is a linear function of the difference between the output voltage signal of the sensor and the threshold signal when the output voltage signal of the sensor is greater than the threshold signal.

20. The first voltage-current conversion circuit, The operational amplifier comprises a first input voltage terminal connected to a threshold signal and a second input terminal connected to the first terminal of a transistor. The IC according to claim 19, wherein the output of the operational amplifier directly drives the second terminal of the transistor, and the third terminal of the transistor operates as the current output terminal of the voltage-to-current conversion circuit.

21. The IC according to claim 19, wherein the first voltage-to-current conversion circuit comprises a MOS transistor.

22. A method for correcting an output signal provided by a magnetoresistive sensor in the presence of an external magnetic field, wherein the correction method is: Determining the deviation of the output signal from the linear response due to the amplitude of the higher-order component signals of the output voltage, The corrected output signal is determined by compensating the output signal of the higher-order component signal so that the linearity error of the corrected output signal is smaller than the linearity error of the output signal. The method further comprises, An integrated circuit (IC) configured to implement a correction method comprising determining an additional signal that is close to or equal to the negative value of the higher-order component signal, wherein determining the corrected output signal comprises compensating the output signal with respect to the higher-order component signal by adding the additional signal to the output signal.

23. The IC according to claim 22, wherein the additional signal is determined by at least one voltage multiplier.

24. The correction output signal Vcorr [Number 16] The system further comprises at least one voltage amplifier (13), Vcorr is the correction output signal, Vout is the output signal, and a1 is the linear coefficient. [Number 17] The IC according to claim 23, wherein a3 is a cubic coefficient and C is a constant.

25. The output signal is input to a comparator, and the output of the comparator is used to activate at least one of at least one multiplexer and one demultiplexer, or any combination of multiplexers and demultiplexers. The IC according to claim 23, further comprising at least one inverting amplifier configured to determine the additional signal determined by at least one voltage multiplier for each polarity of the output signal.

26. The signal offset is added to the output signal, The IC according to claim 23, wherein an input signal corresponding to the sum of the signal offset and the output signal is input to the at least one voltage multiplier.

27. The IC according to claim 26, comprising at least one analog multipurpose unit (AMU) configured to determine the additional signal.

28. The IC according to claim 27, wherein the at least one AMU is based on at least a logarithmic ratio, a logarithm, and an anti-opposite operational amplifier configured to compute the input signal in a power of n, where n is a variable defined by the internal components of the AMU system.

29. Comparator and, A multiplexer and a demultiplexer, The IC according to claim 27, further comprising at least one inverting amplifier, wherein at least one AMU determines the additional signal independently of the polarity of the output signal.

30. The system further comprises a first voltage amplifier with gain G1 and a second voltage amplifier with gain G2. The offset signal is added to the output signal, and an input signal corresponding to the sum of the offset signal and the output signal is input to at least two of the AMUs and the first voltage amplifier. The IC according to claim 27, wherein the corrected output signal is the sum of the output voltage of the AMU and the output voltages of the first and second voltage amplifiers.

31. One of the AMUs is configured to calculate the input signal as a power of 2, The IC according to claim 30, wherein another AMU is configured to calculate the input signal in powers of three.

32. The IC according to claim 22, comprising a digital system (DS) configured to digitally determine the additional signal from the output signal.

33. The IC according to claim 32, wherein the correction method is performed by the DS such that a digital correction output signal is acquired as the final output.

34. The IC according to claim 32, further comprising a digital-to-analog converter (DAC) configured to acquire an analog additional signal from a digitally determined additional signal, such that the correction output signal is obtained by adding the output signal and the analog additional signal.

35. A characteristic evaluation method for deriving common variables for multiple magnetoresistive sensors, A method for correcting an output signal provided by a magnetoresistive sensor in the presence of an external magnetic field, wherein the correction method is: Determining the deviation of the output signal from the linear response due to the amplitude of the higher-order component signals of the output voltage, The method further comprises determining the corrected output signal by compensating the output signal of higher-order component signals such that the linearity error of the corrected output signal is smaller than the linearity error of the output signal, and the method further A characteristic evaluation method in which the common variable is used when performing a correction method, which includes determining an additional signal that is close to or equal to the negative value of the higher-order component signal, wherein determining the corrected output signal includes compensating the output signal with respect to the higher-order component signal by adding the additional signal to the output signal.

36. The output signal Vout is [Number 18] Described by, where H is the external magnetic field, a0 is the offset coefficient, a1 is the first-order coefficient, and the higher-order component signal Vh0 is described by at least the third-order coefficient a3, The correction method includes determining an additional voltage signal corresponding to a negative value of the higher-order component voltage signal, The determination of the corrected output signal includes correcting the output signal to be corrected for the higher-order component signal voltage by adding the additional voltage signal to the output signal, and the characteristic evaluation method is To provide multiple magnetoresistive sensors and measure the output signal for each magnetoresistive sensor, Measured output signal [Number 19] By conforming to this, the offset coefficient a0, the linear coefficient a1, and at least the cubic coefficient a3 are determined, The approximate offset coefficient c0 and the approximate primary coefficient c1 are applied to the measured output signal. [Number 20] It is determined by conforming to the following, The characteristic evaluation method according to claim 35, comprising determining the median value for the determined offset coefficient a0, the linear coefficient a1, at least the cubic coefficient a3, the approximated offset coefficient c0, and the approximated linear coefficient c1.

37. The method according to claim 36, wherein the output signal is measured when the magnetoresistive sensor is exposed to an external magnetic field corresponding to the maximum operating magnetic field range of the magnetoresistive sensor.

38. The method according to claim 36, wherein the plurality of magnetoresistive sensors comprises a lower set of a plurality of magnetoresistive sensors provided within a wafer.

39. When the magnetoresistive sensor is exposed to an external magnetic field corresponding to at least five different magnetic field magnitudes, the output signal is measured, and the magnitudes of the at least five different magnetic fields are A high magnetic field corresponding to the maximum operating magnetic field range of the magnetoresistive sensor, It is provided between the low magnetic field H1, and in the low magnetic field H1, the output signal (Vout) is [Math 21] Within the magnetic field range (-H1, H1) described by the above, the dependence follows a linear relationship, thereby allowing the offset a0 and linear coefficient a1 to be determined by the linear fit of the output voltage Vout, where at least the cubic coefficient a3 is [Number 22] It is derived by the following equation, where Vout_H2 is the output voltage measured in the maximum operating magnetic field range H2, and follows a linear dependence within the magnetic field range (-H1, H1). The measured output signal [Number 23] This is reconstructed from coefficients a0, a1, and a3 previously determined for magnetic fields in the range of -H2 to H2 over any desired magnetic field path. The approximate offset coefficient c0 and the approximate primary coefficient c1 are applied to the measured output signal over the maximum operating magnetic field range. [Number 24] The decision will be made by conforming to the following criteria: The method according to claim 36, comprising determining the median value for the determined offset coefficient a0, the linear coefficient a1, at least the cubic coefficient a3, the approximate offset coefficient c0, and the approximate linear coefficient c1.

Images