The current invention relates to rotary position detection and more specifically to designs for spin-orbit torque (SOT)-based magnetic rotary position sensors.
Various types of sensors have been developed and are being used for rotary position detection. Those sensors include Hall effect sensors and various types of magnetoresistance (MR) sensors. Compared to Hall effect sensors, MR sensors are more accurate and energy-efficient, but they tend to be more expensive due to their complex structures.
Conventional MR sensors include anisotropic magnetoresistance (AMR) sensors, giant magnetoresistance (GMR) or spin-valve (SV) sensors, tunnel magnetoresistance (TMR) sensors, and planar Hall Effect (PHE) sensors. In general, AMR, GMR, SV and TMR sensors operate based on detecting a change in longitudinal resistance when subject to an external magnetic field. PHE sensors operate based on detecting a change in transverse resistance in response to an external field. Compared with the other conventional MR sensors, AMR sensors are typically more robust to electrostatic discharge, and easier to manufacture and use. They also generally have better detective properties despite their low output.
The AMR effect has its origin in spin-orbit coupling (SOC), which results in anisotropic scattering of electrons when they travel through magnetic materials. Materials exhibiting a normal AMR effect show a maximum resistivity when the current is parallel to the magnetization direction (ρ//) and a minimum resistivity when the current is perpendicular to the magnetization direction (ρ⊥). At intermediate angles between the current and magnetization direction, the resistivity of an AMR material is given by ρ(θ)=ρ⊥+(ρ//−ρ⊥)cos2θ, where θ is the angle between the current and the magnetization direction. When the AMR effect is used in magnetic sensing, the magnetization direction is normally set at 45° with respect to the current direction at zero-field so as to maximize the sensitivity, as may be observed by the first derivative of ρ being maximum when ρ=45°. When used in this configuration, the AMR sensor will respond linearly to an external field when the magnitude of the field is small.
The angular dependence of an AMR effect has been exploited for detection of angular positions. In a typical configuration, an in-plane magnetized permanent magnet is attached to a rotation shaft to generate a rotating in-plane field. AMR sensors are placed under the rotating field to detect the magnetic field direction, thereby determining the angular position of the rotational device attached to the shaft.
For each AMR sensor, current may flow between one pair of electrodes of the sensor, and the voltage change caused by an external field may be detected between another pair of electrodes of the sensor. The orientation of one sensor may be set such that it generates a voltage signal proportional to sin (2ϕ), where ϕ is the angle between the current and the external field direction. This is possible when the external field is strong enough to saturate the magnetization in the field direction. The orientation of the other sensor may be 45° away, such that it generates a voltage signal proportional to cos (2ϕ) when subjected to the same rotational field. The output signals evolve two periods when ϕ changes by 360°; therefore, it is only possible to determine rotary position up to 180° unambiguously. In order to resolve 360°, typically an additional sensor must be used to measure the field direction, adding to complexity and manufacturing expense.
A typical GMR sensor consists of two ferromagnetic (FM) layers separated by a non-magnetic spacer and an antiferromagnetic (AFM) layer in contact with one of the FM layers. The thickness of the spacer may be chosen such that there is little exchange coupling between the two FM layers. The magnetization of one of the FM layers that is in direct contact with the AFM layer is pinned by the latter, and thus this FM layer is commonly called a “pinned” layer. The magnetization of the other FM layer is free to rotate to respond to an external field, and thus it is commonly called a “free” layer. The exchange bias between the AFM and FM sets the direction between the magnetizations of the free and pinned layer at 90° at zero-field. This is to ensure that the sensor will respond to external field linearly. When being used, a constant current is applied to the sensor and the voltage change caused by external field is detected. The output signal is proportional to sin(ϕ), where ϕ is the angle between the magnetization of the free layer and that of the pinned layer.
A TMR sensor typically has the same structure as that of a GMR sensor. The main difference is that, in the case of TMR, the non-magnetic layer is replaced by an insulator. In addition, instead of flowing in-plane, current flows vertically with respect to the layers. Again, the output signal is proportional to sin(ϕ), where ϕ is the angle between the magnetization of the free layer and that of the pinned layer.
From the angle dependence of output signal, it is apparent that both GMR and TMR sensors can be used for rotary position detection as long as the external field is able to saturate the magnetization of the free layer without disturbing the magnetization of the reference layer. As with the case of rotary position detection using AMR sensors, multiple GMR or TMR sensors may be placed under the rotating field to detect the magnetic field direction, thereby determining the angular position of a rotational device. With GMR and TMR sensors, it may be possible to determine the angle unambiguously up to 360°, without the need for additional sensors to determine field direction. However, additional processes are typically required to set and align the pinning directions of individual sensors, which significantly increases manufacturing and implementation costs for GMR and TMR sensors. Further, the magnetization of the reference layer may be changed by thermal effect after the sensor has been used for some time in a hush environment.
Accordingly, there remains a need for a magnetic rotary position sensor that may achieve high detection accuracy while being relatively simple and cost effective to manufacture.
In example embodiments, a SOT-based magnetic rotary position sensor includes two spin Hall anomalous Hall effect (SHAHE) sensors, two spin Hall magnetoresistance (SMR) sensors or two unidirectional spin Hall magnetoresistance (USMR) sensors. In embodiments using SHAHE sensors, the sensors may be structured as Hall crosses formed from a film stack including a FM/heavy metal (HM) bi-layer or other SOT-generating layers. The current axes of the Hall crosses are orthogonally aligned. In embodiments using SMR or USMR sensors, the sensors may be structured as Wheatstone bridges including four SMR or USMR sensing elements each formed from a film stack including a FM/HM bi-layer or other SOT-generating layers. The field sensing axes of the Wheatstone bridges are orthogonally aligned. In such an arrangement of two SHAHE, SMR or USMR sensors, one of the sensors may produce a sine-like waveform and the other a cosine-like waveform, when the sensors are subject to an external rotating magnetic field parallel (in-plane) to the layers (e.g., caused by a magnet attached to a shaft of a rotational device, such as a motor). The sine-like waveform and the cosine-like waveform complete a single period when the in-plane magnetic field rotates by 360°. The angle of the rotational device can be readily calculated (e.g., by circuitry), by applying appropriate algorithms to the sine-like and cosine-like waveforms. Such a SOT-based magnetic rotary position sensor may achieve high detection accuracy while being more efficient and cost effective to manufacture and implement than conventional sensor designs, among other advantages.
It should be understood that a variety of additional features and alternative embodiments may be implemented other than those discussed in this Summary. This Summary is intended simply as a brief introduction to the reader, and does not indicate or imply that the examples mentioned herein cover all aspects of the disclosure, or are necessary or essential aspects of the disclosure.
The description below refers to the accompanying drawings of example embodiments, of which:
Consider first an embodiment where the sensors 120, 130 are SHAHE sensors that each include a Hall cross, and the current axes of the Hall crosses are orthogonally aligned.
Each Hall cross 210, 220 includes a stack of ultrathin films. Such films may have a number of different configurations, depending on the implementation.
The film stack (e.g., FM/HM bilayer structure) is capable of generating SOT. SOT is a promising mechanism for magnetization switching and related applications. It is generally accepted that two types of torques are present in a FM/HM bilayer: one commonly referred to as the field-like (FL) torque and the other commonly referred to as the (anti)damping-like (DL) torque. The two types of torques can be modelled by {right arrow over (T)}DL=−τDL{right arrow over (m)}×[m{right arrow over (m)}×({right arrow over (j)}×{right arrow over (z)})] and {right arrow over (T)}FL=−τFL{right arrow over (m)}×({right arrow over (j)}×{right arrow over (z)}), respectively, where {right arrow over (m)} is the magnetization direction, j is the in-plane current density, {right arrow over (z)} is the interface normal, and τFL and τDL are the magnitude of the FL and DL torques, respectively. If {right arrow over (m)}. does not change significantly, the two toques can be expressed in the form of {right arrow over (M)}×{right arrow over (H)}eff, where {right arrow over (H)}eff is an effective field. Following this notion, the FL effective field ({right arrow over (H)}FL) is in the direction of −{right arrow over (j)}×{right arrow over (z)}, whereas the DL effective field ({right arrow over (H)}DL) is in the direction of −{right arrow over (m)}×({right arrow over (j)}×{right arrow over (z)}) (note: the sign can be different depending on the stacking order of HM and FM with respect to the coordinate axes).
The change of AHE and PHE signal induced by the SOT effective fields can be derived analytically by assuming (1) the effective field is much smaller than the externally applied field {right arrow over (H)}ex and (2) both the in-plane magnetic and shape anisotropy of the FM layer can be neglected. The former allows one to evaluate the influence of effective field using first order approximation, whereas the latter makes it possible to decompose the effective field into different components and calculate their effects on magnetization separately. Based on these assumptions, the Hall voltage can be written as Vxy=Vxy({right arrow over (H)}ex)+ΔVxy({right arrow over (H)}l), where {right arrow over (H)}I={right arrow over (H)}FL+{right arrow over (H)}DL+{right arrow over (H)}Oe is the sum of SOT effective field and Oersted field {right arrow over (H)}Oe. When the applied external field is in-plane, and both in-plane magnetic and shape anisotropy can be neglected, one may assume θm≈θH=π/2 and φm≈φH. Under these approximations, the differential output signal from pulsed current measurement is given by:
where VAHE0 and VPHE0 are the half peak-to-peak voltage of AHE and PHE, respectively, φH is the in-plane field angle with respect to x axis, and Hd is the demagnetizing field.
It is apparent from Eq. (1) that when the PHE is negligibly small, a Hall cross including a FM/HM bilayer structure functions as an angular position sensor as ΔVxy({right arrow over (H)}i) ∝cosφH. The advantage of using a SOT Hall device is that, any offset in the device can be readily compensated by measuring the AHE first using a positive current (+I) followed by a negative current of same magnitude (−I) and then adding up the two measurement results to generate the output signal. This is possible because VAHE is an even function, whereas the offset voltage is an odd function of the driving current. The output signal is simply double of that of the single measurement.
In a second embodiment, each sensor 120, 130 may be a SMR sensor that is structured as a Wheatstone bridge.
The film stack of each SMR sensing element 802-809 may include a FM/HM bilayer structure. The FM layer may include a material such as Co, Fe, Ni, CoFeB, Gd, YIG, a ferrite, and/or alloys comprising Co, Fe, Ni, CoFeB or Gd. The HM material may include a material such as Pt, Pd, Ta, W, Pb, Nb, a topological insulator, a TMD and/or a Weyl metal or semimetal. When an AC current I=I0 sinωt passes through a sensing element 802-809 including a FM/HM bilayer, the voltage across the two ends of the sensing element can be derived as follows:
The time-average of V, or DC component, is given by:
where I0 is the amplitude of applied AC current, ω is the angular frequency, α is the ratio of SOT effective field over applied current, HFL=αI0 sinωt, ΔR is magnetoresistance, Ro is resistance of sensing element, HK is the uniaxial anisotropy field, HD is the shape anisotropy field, Hy is the applied field, and θ is the angle between magnetization (M) and easy axis direction. θ can be determined by the energy minimization method and is given by
It is clear from Eq. (3) that, although the sense current is an AC current, the output signal has a DC component that is proportional to the external magnetic field in the y-direction. This means that the sensor exhibits a linear response to external field without the requirement that θ must be 45° at Hy=0. This is in sharp contrast to DC biasing in which θ must be set at 45° at Hy=0 in order to ensure a linear response of the sensor. Alternatively, the sensor output can also be obtained by detecting the 2nd harmonic of the signal given by Eq. (1), for example using a lock-in technique.
In the case of a Wheatstone bridge 710, 720, when an AC current I=lo sinωt passes through two electrodes, the voltage across the other two electrodes is derived as:
where ΔR0 is the offset resistance caused by non-identical SMR sensing elements due to manufacturing processes, and the remaining parameters are the same as those used in Eq. (2) and (3). The time-average of V, or DC component, is given by:
Again, it is clear from Eq. (5) that, although the sense current is an AC current, the output signal has a DC component which is proportional to the external magnetic field. This means that the sensor exhibits a linear response to external field without the requirement that θ must be 45° at Hy=0. Compared to the case of a single sensing element, a Wheatstone bridge configuration leads to a sensor with much smaller AC noise as the large AC signal due to the resistance of each sensing element has been cancelled out. It has also has a smaller DC offset and thermal drift. The DC offset, if any, can be further reduced by adding a DC offset to sense/bias current. Alternatively, lock-in detection may be employed to detect the 2nd harmonic which is also proportional to the external field.
Since the output voltage is proportional to Hy when Hy is small, when a magnetic field with constant amplitude is rotation in the xy plane, a sine-like waveform is generated from the first Wheatstone bridge 710 and a cosine-like waveform is generated from the second Wheatstone bridge 720.
Once the outputs of the first Wheatstone bridge 710 and second Wheatstone bridge 720 are obtained, circuitry may calculate the field angle by applying an acctan 2 function, for example as arctan 2 (output of Wheatstone bridge 710/Wheatstone bridge 720). Pre-calibration may be performed to compensate for deviation (if any) from ideal sine or cosine curves so as to reduce the angle error.
In some implementations, a magnetic shield (not shown) may be used to reduce the field generated by the magnet 150, such that the field will be within the operation range of the SMR sensors. By doing so, the effect of environmental field can be effectively suppressed.
In a third embodiment, each sensor 120, 130 may be a USMR sensor that includes a Wheatstone bridge formed from for USMR sensing elements. Again, the Wheatstone bridges each detect the direction of the magnetic field generated by a magnet attached to the shaft, coupled to a rotational device, for example, a motor, similar to as in
The film stack of each USMR sensing element may include a FM/HM bilayer structure. The FM layer may include a material such as Co, Fe, Ni, CoFeB, Gd, YIG, a ferrite, and/or alloys comprising Co, Fe, Ni, CoFeB or Gd. The HM material may include a material such as Pt, Pd, Ta, W, Pb, Nb, a topological insulator, a TMD and/or a Weyl metal or semimetal. When charge current flows through the bilayer, spin current is generated in the HM layer due to spin Hall effect. The spin current diffuses into the FM layer, causing an additional interfacial resistance that is sensitive to the magnetization direction of the FM layer with respect to the spin polarization direction. This gives a USMR, which is a sinusoidal function of the angle between the magnetization direction of the FM layer and the spin polarization direction of the spin current.
In alternative implementations, each USMR sensing element may be formed from a single material with unique spin texture on the surface, such as topological insulators or Weyl metal or semimetal. In still other alternative implementations, each USMR may be an antiferromagnet.
In summary, various embodiments of the present disclosure describe a SOT-based magnetic rotary position sensor. Depending on the embodiment, the sensor may include SHAHE Hall crosses, Wheatstone bridges formed from SMR sensing elements or Wheatstone bridges formed from USMR sensing elements. It should be understood that numerous adaptations and modifications may be made to the above-discussed embodiments without departing from the disclosures intended spirit and scope.
For example, while it is discussed above that a SHAHE Hall cross, SMR sensing elements or USMR sensing elements may include a FM/HM bilayer that generates SOT, it should be understood that other SOT-generating layer structures may alternative be used. For example, a FM/normal-metal (NM)/HM trilayer may be instead used. In such an alternative embodiment, the NM layer may be, for example, silver (Ag), copper (Cu), gold (Au) or another material, and the HM layer may be, for example, bismuth (Bi), bismuth oxide, a topological insulator, or another material. In such an alternative embodiment, the spin orbit torque effect may be generated from both the spin Hall and/or Rashba-Edelstein effect. In other alternative embodiments, an antiferromagnet (AFM)/HM bilayer, a FM/AFM/HM trilayer, an AFM/HM/FM trilayer, an AFM/FM/HM trilayer, or a HM/AFM/FM trilayer may be used instead of a FM/HM bilayer. The AFM may include a material such as FeMn, IrMn, NiFe, PtMn, NiMn, PtNiMn, Mn, Cr, NiO, CoO or CuMnAs. The FM may include a material such as Co, Fe, Ni, CoFeB Gd, YIG, a ferrite, or alloys comprising Co, Fe, Ni, CoFeB or Gd. The HM may include a material such as Pt, Pd, Ta, W, Pb, Nb, a topological insulator, TMD or a Weyl metal or semimetal.
In general, it should be remembered that the various elements described above may be made from differing materials, implemented in different combinations or otherwise formed or used differently. Example embodiments are not necessarily mutually exclusive as some may be combined with one or more embodiments to form new example embodiments. Figures are not drawn to scale and relative relationships in size may be exaggerated for clarity in presentation. What is claimed is:
Number | Date | Country | Kind |
---|---|---|---|
10201805003Q | Jun 2018 | SG | national |
The present application claims priority to Singapore Patent Application No. 10201805003Q, filed by Applicant National University of Singapore on Jun. 12, 2018, the contents of which are incorporated by reference herein in their entirety.