rus
Issue #3-4/2023
V.V.Polevikov, E.O.Litvinenko
MODERN METHODS FOR MONITORING ELECTRICAL CURRENTS USING PHYSICAL VALUES SENSORS
MODERN METHODS FOR MONITORING ELECTRICAL CURRENTS USING PHYSICAL VALUES SENSORS
DOI: https://doi.org/10.22184/1993-8578.2023.16.3-4.220.230
Magnetic field sensors are very important for many sensor-based measurement applications. Therefore, measuring current values comes down to a choice between an AMR sensor and a Hall sensor. This paper presents the main advantages and disadvantages of each of the two types of sensors, methods for optimising their main characteristics, as well as the results achieved by Zelenograd Nanotechnology Centre in mastering and producing magnetic transducers.
Magnetic field sensors are very important for many sensor-based measurement applications. Therefore, measuring current values comes down to a choice between an AMR sensor and a Hall sensor. This paper presents the main advantages and disadvantages of each of the two types of sensors, methods for optimising their main characteristics, as well as the results achieved by Zelenograd Nanotechnology Centre in mastering and producing magnetic transducers.
Теги: current sensor detectors hall sensor magnetoresistors амr датчики датчики тока магниторезисторы холл
INTRODUCTION
The optimal solution for monitoring electrical current is the use of small, reliable magnetic field to electrical signal conversion devices called sensors. The basic tools for this are classic Hall sensors and devices on anisotropic magnetoresistive films - AMR converters.
In all applications of physical values sensors in signal measurement, there is a set of parameters that must be monitored for the stable operation of the device where the signal from that sensor is applied. This paper presents the main advantages and disadvantages of each of the two types of sensors presented, methods for optimising their characteristics, such as zero offset voltage and 1/f flicker noise, and the results achieved by Zelenograd Nanotechnology Centre in developing and producing magnetic transducers.
Monitoring currents with physical values sensors
Magnetic field sensors are widely used in many applications such as industry, medicine, mobile technology and especially automotive electronics, which is the leading (over 50% of the total market volume) area of application [1]. They provide convenient, non-contact, robust and reliable operating devices compared to many other sensors. Two types of sensors – Hall-effect sensors and so-called AMR sensors – dominate in these industries for their use in measuring systems.
The Hall effect in semiconductors is manifested when it is placed in a magnetic field. The Hall voltage will then be generated perpendicular to both the current and the field. Figure 1 shows a thin sheet of semiconductor material (a Hall element) when a current is passed. The output connections are perpendicular to the current direction. If magnetic field is absent (a), the current distribution is uniform and there is no potential difference at the output.
When magnetic field is perpendicular (b), a Lorentz force acts. This force disturbs the current distribution, resulting in a potential difference (voltage) across the coils. This voltage is the Hall voltage (VH) [2].
The principle of AMR sensors operation is based on the effect of a magnetic film changing in a magnetic field, where material resistance tends to change under external magnetic fields influence. In anisotropic magnetoresistance (AMR) the resistance changes (ΔR) depending on the angle θ between electric current and metal magnetisation (Fig.2). As in the case of Hall effect sensors, the main cause of magnetoresistance is the Lorentz force, which causes electrons to move along a curved path between collisions [3].
Hall effect sensors respond to magnetic fields perpendicular to the sensor. On the other hand, AMR sensors respond to planar fields and react to both magnetic poles. Therefore, AMR sensor has several layout options available to detect planar magnetic fields and as a result it offers a wide design flexibility. Hall effect sensors, in opposite, typically perform accurate magnetic field detection in perpendicular direction, and because it is recommended that the magnet be located directly above the Hall effect sensor and at close range, it has limited design flexibility (Fig.3) [4].
Hall sensors have a higher linearity (Hall EMF is linearly dependent on the applied magnetic field) than AMR sensors. This linearity is evident in the range up to a few thousand Gs. The nature of AMR sensors is different, as the linear section of their transfer characteristic is short, and saturation occurs at induction of a few tens (maximum hundreds) of Gs. However, AMR sensors sensitivity is much higher than that of Hall effect sensors, which allows them to detect even the weakest magnetic fields (from 1mT).
Another difference between AMR and Hall sensors is that they can be manufactured in bulk on silicon wafers and mounted in commercially available integrated circuit packages. This allows magnetic sensors to be automatically assembled with other circuit and system components. Hall sensors are often also integrated directly into a wafer with necessary amplifier stage to generate signals.
However, when using with both Hall and AMR sensors, you must be aware of the problems that can occur when using them. The zero offset voltage in sensors is voltage that can be measured at the sensor output if magnetic field is absent. It arises due to imperfect sensor design and, just like noise in sensor structures, can lead to measurement errors. Therefore, in order to accurately measure electrical currents with physical values sensors, these parameters must be taken into account and compensated for.
Offset voltage and 1/f flicker noise correction for Hall sensors
In order to reduce the output signal bias, there are several methods. The first method involves a correction circuit at the input of the amplifier stage. A voltage equal in magnitude to the offset voltage, but opposite in sign, will be applied to the input of the amplifier in addition to the useful signal. This method will thus compensate the voltages and minimise the error (Fig.4). By varying the number of trim inputs and the resistors size, accuracy and magnitude of the offset correction can be varied.
The second method uses an even number of Hall elements connected orthogonally. The bias voltage depends on the Hall sensor geometry, whereas the Hall voltage and, hence, sensitivity are independent functions. As a result, the outputs of the devices can be connected in such a way that the corresponding Hall voltages are averaged and the bias voltages are compensated [5]. Thus, Fig.5 and 6 show that using this method makes it is possible to reduce the bias by almost half.
With the third method (chopper stabilisation method) low frequency errors such as 1/f noise and drift will be modulated and filtered out along with the offset [5]. This can be seen in Fig.7, which depicts the chopper stabilisation method in the frequency domain. Firstly, the signal is modulated and noise and bias are superimposed on this modulated signal (Fig.7, b). After amplification and modulation it is necessary to demodulate the signal back to DC and low frequency noise and offset are modulated at the clock frequency, appearing as a ripple at the amplifier output (Fig.7, c). A low pass filter is then used to remove the modulated offset and 1/f noise, resulting in only low frequency signal with no offset or 1/f noise (Fig 7, d) [6].
To remove the 1/f noise completely, the chopper frequency must be higher than angular frequency of the 1/f noise. This method can successfully remove error and noise from the amplifiers that are the basis of the processing channel. However, this method only removes error and low-frequency noise at the amplifier, but does not do the same for the transducer output.
The fourth method involves the use of a current rotator circuit, which is an alternative method of obtaining a signal from the probe (current rather than voltage mode of operation). During each phase Fs<1:4> the bias current Ibias is delivered to two consecutive arms; for a zero magnetic field it is divided into two equal parts, assuming that the voltages on the remaining two pins are equal. The output and supply terminals of each Hall plate are periodically reversed, so that polarity of the input current is reversed in each state, while the bias appears on the output terminals. At the end of a complete rotation operation, the average offset value becomes zero (Fig.8) [7].
Since the current rotator implementation involves only the Hall elements themselves and CMOS switches, this circuit will not take up much space on-chip and it also successfully combats 1/f noise and sensor bias. The disadvantages are the output signal attenuation, which can be eliminated by using an amplifier with a higher Ку.
Offset voltage and 1/f flicker noise correction for AMR sensors
An effective way to reduce temperature drift of the bias voltage by using a stabilising AC field has been proposed by Tumansky (1984). It uses a special planar coil to generate stabilising field. Figure 9 shows this method using reset (RESET) and set (SET) signals.
The transfer characteristic of the sensor shows slope inversion and the crossover point at zero offset voltage. To measure the applied field H, the sensor is first activated by a SET setting signal (Fig.10).
Once the Vset voltage has been established, it is read out and recorded. The reset signal is then applied and the Vreset voltage is read. This gives us two values:
VSET = SHx + VOFFSET (1)
VRESET = –SHx + VOFFSET. (2)
The difference between two values is only proportional to the field magnitude and the bias voltage is not taken into account [8]:
VSET – VRESET = 2SHx. (3)
In addition, advantage of performing modulation methods using switching pulses is to reduce the 1/f noise within the desired bandwidth [9]. In addition to the SET/RESET coil there is possibility of using an OFFSET coil which will compensate for any external magnetic field if required.
JSC “Zelenograd Nanotechnology Center” products
Zelenograd Nanotechnology Centre developed a technology for producing magnetoresistive sensitive elements (MRSE), which includes applying magnetoresistive films with a magnetoresistive effect of 2% to 3.5%, photolithography to form on their basis a system of multiple-layer conductors, and assemble them into packages.
This led to development and manufacture of circuits such as 5202HX01H4 and 1382HX065. The first one is designed for navigation, guidance and orientation in space, object movement detection and anomaly detection. MRSE of the first circuit is based on 80Ni20Fe film, with magnetoresistance effect value ~ 2.2%. The MRSE of this chip has characteristics similar to Honeywell’s HMC1022 sensor.
The second microcircuit is designed for tracking systems of rudder angle control of aircrafts and tracking system drives parameters. Second MRSE circuit is made on 74Ni10Fe16Co film with magnetoresistance effect value ~ 3.5%. As a result, this circuit achieves 100 mV output at 5 V supply voltage with high degree of output signal synchronization. The chip is an analogue of the world’s leading microcircuits in its characteristics.
In addition, in ZNC JSC was developed the design and manufacturing base and has mastered production of a range of Hall and AMR sensor microcircuits, including not only microcircuits for monitoring electrical currents of large magnitudes, but also for measuring rotation angle, processing nonius signals, encoders, angular position, and for use in special-purpose devices, etc.
Microcircuit K1382NU01A5 (Fig.12) is designed to convert the measured value of magnetic field of conductor with current into the output signals of standard interfaces of proximity current sensor. An external magnetoresistive or internal Hall effect sensor can be used as a current proximity sensor. The sensor is placed over the conductor of the printed circuit board. The chip amplifies the sensor signals, converts them and outputs the signal as standard interfaces. The chip contains a built-in system for correcting magnetosensitive element temperature dependence.
CONCLUSIONS
In conclusion, AMR and Hall sensors are virtually indispensable tools in magnetic field applications due to their relative technological simplicity, reliability and low price. The choice between these two devices is determined solely by the customer’s specifications, virtually all of the serious drawbacks of both sensors can be removed using the methods described in this paper.
PEER REVIEW INFO
Editorial board thanks the anonymous reviewer(s) for their contribution to the peer review of this work. It is also grateful for their consent to publish papers on the journal’s website and SEL eLibrary eLIBRARY.RU.
Declaration of Competing Interest. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
The optimal solution for monitoring electrical current is the use of small, reliable magnetic field to electrical signal conversion devices called sensors. The basic tools for this are classic Hall sensors and devices on anisotropic magnetoresistive films - AMR converters.
In all applications of physical values sensors in signal measurement, there is a set of parameters that must be monitored for the stable operation of the device where the signal from that sensor is applied. This paper presents the main advantages and disadvantages of each of the two types of sensors presented, methods for optimising their characteristics, such as zero offset voltage and 1/f flicker noise, and the results achieved by Zelenograd Nanotechnology Centre in developing and producing magnetic transducers.
Monitoring currents with physical values sensors
Magnetic field sensors are widely used in many applications such as industry, medicine, mobile technology and especially automotive electronics, which is the leading (over 50% of the total market volume) area of application [1]. They provide convenient, non-contact, robust and reliable operating devices compared to many other sensors. Two types of sensors – Hall-effect sensors and so-called AMR sensors – dominate in these industries for their use in measuring systems.
The Hall effect in semiconductors is manifested when it is placed in a magnetic field. The Hall voltage will then be generated perpendicular to both the current and the field. Figure 1 shows a thin sheet of semiconductor material (a Hall element) when a current is passed. The output connections are perpendicular to the current direction. If magnetic field is absent (a), the current distribution is uniform and there is no potential difference at the output.
When magnetic field is perpendicular (b), a Lorentz force acts. This force disturbs the current distribution, resulting in a potential difference (voltage) across the coils. This voltage is the Hall voltage (VH) [2].
The principle of AMR sensors operation is based on the effect of a magnetic film changing in a magnetic field, where material resistance tends to change under external magnetic fields influence. In anisotropic magnetoresistance (AMR) the resistance changes (ΔR) depending on the angle θ between electric current and metal magnetisation (Fig.2). As in the case of Hall effect sensors, the main cause of magnetoresistance is the Lorentz force, which causes electrons to move along a curved path between collisions [3].
Hall effect sensors respond to magnetic fields perpendicular to the sensor. On the other hand, AMR sensors respond to planar fields and react to both magnetic poles. Therefore, AMR sensor has several layout options available to detect planar magnetic fields and as a result it offers a wide design flexibility. Hall effect sensors, in opposite, typically perform accurate magnetic field detection in perpendicular direction, and because it is recommended that the magnet be located directly above the Hall effect sensor and at close range, it has limited design flexibility (Fig.3) [4].
Hall sensors have a higher linearity (Hall EMF is linearly dependent on the applied magnetic field) than AMR sensors. This linearity is evident in the range up to a few thousand Gs. The nature of AMR sensors is different, as the linear section of their transfer characteristic is short, and saturation occurs at induction of a few tens (maximum hundreds) of Gs. However, AMR sensors sensitivity is much higher than that of Hall effect sensors, which allows them to detect even the weakest magnetic fields (from 1mT).
Another difference between AMR and Hall sensors is that they can be manufactured in bulk on silicon wafers and mounted in commercially available integrated circuit packages. This allows magnetic sensors to be automatically assembled with other circuit and system components. Hall sensors are often also integrated directly into a wafer with necessary amplifier stage to generate signals.
However, when using with both Hall and AMR sensors, you must be aware of the problems that can occur when using them. The zero offset voltage in sensors is voltage that can be measured at the sensor output if magnetic field is absent. It arises due to imperfect sensor design and, just like noise in sensor structures, can lead to measurement errors. Therefore, in order to accurately measure electrical currents with physical values sensors, these parameters must be taken into account and compensated for.
Offset voltage and 1/f flicker noise correction for Hall sensors
In order to reduce the output signal bias, there are several methods. The first method involves a correction circuit at the input of the amplifier stage. A voltage equal in magnitude to the offset voltage, but opposite in sign, will be applied to the input of the amplifier in addition to the useful signal. This method will thus compensate the voltages and minimise the error (Fig.4). By varying the number of trim inputs and the resistors size, accuracy and magnitude of the offset correction can be varied.
The second method uses an even number of Hall elements connected orthogonally. The bias voltage depends on the Hall sensor geometry, whereas the Hall voltage and, hence, sensitivity are independent functions. As a result, the outputs of the devices can be connected in such a way that the corresponding Hall voltages are averaged and the bias voltages are compensated [5]. Thus, Fig.5 and 6 show that using this method makes it is possible to reduce the bias by almost half.
With the third method (chopper stabilisation method) low frequency errors such as 1/f noise and drift will be modulated and filtered out along with the offset [5]. This can be seen in Fig.7, which depicts the chopper stabilisation method in the frequency domain. Firstly, the signal is modulated and noise and bias are superimposed on this modulated signal (Fig.7, b). After amplification and modulation it is necessary to demodulate the signal back to DC and low frequency noise and offset are modulated at the clock frequency, appearing as a ripple at the amplifier output (Fig.7, c). A low pass filter is then used to remove the modulated offset and 1/f noise, resulting in only low frequency signal with no offset or 1/f noise (Fig 7, d) [6].
To remove the 1/f noise completely, the chopper frequency must be higher than angular frequency of the 1/f noise. This method can successfully remove error and noise from the amplifiers that are the basis of the processing channel. However, this method only removes error and low-frequency noise at the amplifier, but does not do the same for the transducer output.
The fourth method involves the use of a current rotator circuit, which is an alternative method of obtaining a signal from the probe (current rather than voltage mode of operation). During each phase Fs<1:4> the bias current Ibias is delivered to two consecutive arms; for a zero magnetic field it is divided into two equal parts, assuming that the voltages on the remaining two pins are equal. The output and supply terminals of each Hall plate are periodically reversed, so that polarity of the input current is reversed in each state, while the bias appears on the output terminals. At the end of a complete rotation operation, the average offset value becomes zero (Fig.8) [7].
Since the current rotator implementation involves only the Hall elements themselves and CMOS switches, this circuit will not take up much space on-chip and it also successfully combats 1/f noise and sensor bias. The disadvantages are the output signal attenuation, which can be eliminated by using an amplifier with a higher Ку.
Offset voltage and 1/f flicker noise correction for AMR sensors
An effective way to reduce temperature drift of the bias voltage by using a stabilising AC field has been proposed by Tumansky (1984). It uses a special planar coil to generate stabilising field. Figure 9 shows this method using reset (RESET) and set (SET) signals.
The transfer characteristic of the sensor shows slope inversion and the crossover point at zero offset voltage. To measure the applied field H, the sensor is first activated by a SET setting signal (Fig.10).
Once the Vset voltage has been established, it is read out and recorded. The reset signal is then applied and the Vreset voltage is read. This gives us two values:
VSET = SHx + VOFFSET (1)
VRESET = –SHx + VOFFSET. (2)
The difference between two values is only proportional to the field magnitude and the bias voltage is not taken into account [8]:
VSET – VRESET = 2SHx. (3)
In addition, advantage of performing modulation methods using switching pulses is to reduce the 1/f noise within the desired bandwidth [9]. In addition to the SET/RESET coil there is possibility of using an OFFSET coil which will compensate for any external magnetic field if required.
JSC “Zelenograd Nanotechnology Center” products
Zelenograd Nanotechnology Centre developed a technology for producing magnetoresistive sensitive elements (MRSE), which includes applying magnetoresistive films with a magnetoresistive effect of 2% to 3.5%, photolithography to form on their basis a system of multiple-layer conductors, and assemble them into packages.
This led to development and manufacture of circuits such as 5202HX01H4 and 1382HX065. The first one is designed for navigation, guidance and orientation in space, object movement detection and anomaly detection. MRSE of the first circuit is based on 80Ni20Fe film, with magnetoresistance effect value ~ 2.2%. The MRSE of this chip has characteristics similar to Honeywell’s HMC1022 sensor.
The second microcircuit is designed for tracking systems of rudder angle control of aircrafts and tracking system drives parameters. Second MRSE circuit is made on 74Ni10Fe16Co film with magnetoresistance effect value ~ 3.5%. As a result, this circuit achieves 100 mV output at 5 V supply voltage with high degree of output signal synchronization. The chip is an analogue of the world’s leading microcircuits in its characteristics.
In addition, in ZNC JSC was developed the design and manufacturing base and has mastered production of a range of Hall and AMR sensor microcircuits, including not only microcircuits for monitoring electrical currents of large magnitudes, but also for measuring rotation angle, processing nonius signals, encoders, angular position, and for use in special-purpose devices, etc.
Microcircuit K1382NU01A5 (Fig.12) is designed to convert the measured value of magnetic field of conductor with current into the output signals of standard interfaces of proximity current sensor. An external magnetoresistive or internal Hall effect sensor can be used as a current proximity sensor. The sensor is placed over the conductor of the printed circuit board. The chip amplifies the sensor signals, converts them and outputs the signal as standard interfaces. The chip contains a built-in system for correcting magnetosensitive element temperature dependence.
CONCLUSIONS
In conclusion, AMR and Hall sensors are virtually indispensable tools in magnetic field applications due to their relative technological simplicity, reliability and low price. The choice between these two devices is determined solely by the customer’s specifications, virtually all of the serious drawbacks of both sensors can be removed using the methods described in this paper.
PEER REVIEW INFO
Editorial board thanks the anonymous reviewer(s) for their contribution to the peer review of this work. It is also grateful for their consent to publish papers on the journal’s website and SEL eLibrary eLIBRARY.RU.
Declaration of Competing Interest. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Readers feedback
