Abstract Delta-E effect sensors are based on magnetoelectric resonators that detune in a magnetic field due to the delta-E effect of the magnetostrictive material. In recent years, such sensors have shown the potential to detect small amplitude and low-frequency magnetic fields. Yet, they all require external magnetic bias fields for optimal operation, which is highly detrimental to their application. Here, we solve this problem by combining the delta-E effect with exchange biased multilayers and operate the resonator in a low-loss torsion mode. It is comprehensively analyzed experimentally and theoretically using various kinds of models. Due to the exchange bias, no external magnetic bias fields are required, but still low detection limits down to $${{\text{350 pT}} \mathord{\left/ {\vphantom {{\text{350 pT}} {\sqrt {{\text{Hz}}} }}} \right. \kern-\nulldelimiterspace} {\sqrt {{\text{Hz}}} }}$$ 350 pT / Hz at 25 Hz are achieved. The potential of this concept is demonstrated with a new operating scheme that permits simultaneous measurement and localization, which is especially desirable for typical biomedical inverse solution problems. The sensor is localized with a minimum spatial resolution of 1 cm while measuring a low-frequency magnetic test signal that can be well reconstructed. Overall, we demonstrate that this class of magnetic field sensors is a significant step towards first biomedical applications and compact large number sensor arrays.

Despite obvious advantages over their electric counterparts, magnetic measurements are still performed very rarely for medical diagnosis. The high operating costs of established systems with a sufficient limit of detection, mainly based on super-conducting quantum interference devices, prevent the spread of magnetic diagnostic procedures. Consequently, the demand for alternative low-cost magnetic sensor systems is large. Several suitable, uncooled sensor concepts exist, but their usability-especially, when operated outside magnetically shielded chambers-requires improvement. This paper highlights basic, real-time signal processing concepts, and implementations to improve the signal quality for different biomedical applications with a focus on cardiology in unshielded measurement environments. Several processing steps in the digital domain are described, such as noise reduction (by means of cancellation and suppression), adaptive sensor signal combination, and adaptive signal averaging. Finally, basic cardiologic feature extraction methods are performed using uncooled magnetometers in combination with the described signal enhancement stages.

Recently, there has been much interest in magnetoelectric magnetic field sensors utilizing the delta-E effect. Such sensors are fully integrable and combine the advantages of high sensitivity at low frequencies with broad bandwidth. Here, we report the influence of the quality factor Q on the signal-to-noise ratio of magnetoelectric magnetic field sensors utilizing the delta-E effect. The sensor consists of a silicon cantilever covered by a magnetostrictive and a piezoelectric thin film. The magnetization-dependent elasticity of the magnetostrictive film leads to detuning of the sensor's resonance, which is excited and read out via the piezoelectric layer. The signal-to-noise ratio is experimentally analyzed as a function of the quality factor, the excitation amplitude and the signal frequency. The results are compared with a signal and noise model to describe general tendencies. The model demonstrates that, in contrast to the conventional direct operation of magnetoelectric sensors, an improvement in the limit of detection proportional to Q3/2 can be achieved if thermal-mechanical noise is dominant. The relationship still holds for frequencies far away from the resonance frequency. This reveals the potential for improving the limit of detection significantly by increasing the quality factor, if magnetic and electronic noise can be suppressed.

To enable the measurement of low-frequency magnetic signals with cantilever type thin-film magnetoelectric sensors, magnetic frequency conversion transfers the frequency of the desired signal into the mechanical resonance of the cantilever. The system electronics for the realization of this approach and the approach itself introduce additional noise sources as compared with direct detection, which lowers the limit of detection. In this paper, the magnetic frequency conversion noise sources are reviewed, discussed, and evaluated for our setup. The model for the nonlinear transfer process is implemented in the time domain. This enables the consideration of the pump noise in a noise equivalent circuit. For the sensor type under investigation, the dominant noise near its optimal working point originates from the pump source. If the noise of the pump can be decreased and magnetic excess noise is not dominant, the noise limit is the thermal-mechanical noise of the sensor. The implementation of a filter after the excitation source decreases the limit of detection to 60 pT/root Hz at 10 Hz.

We present a comprehensive study of a magnetic sensor system that benefits from a new technique to substantially increase the magnetoelastic coupling of surface acoustic waves (SAW). The device uses shear horizontal acoustic surface waves that are guided by a fused silica layer with an amorphous magnetostrictive FeCoSiB thin film on top. The velocity of these so-called Love waves follows the magnetoelastically-induced changes of the shear modulus according to the magnetic field present. The SAW sensor is operated in a delay line configuration at approximately 150 MHz and translates the magnetic field to a time delay and a related phase shift. The fundamentals of this sensor concept are motivated by magnetic and mechanical simulations. They are experimentally verified using customized low-noise readout electronics. With an extremely low magnetic noise level of ≈100 pT/ √(Hz) , a bandwidth of 50 kHz and a dynamic range of 120 dB, this magnetic field sensor system shows outstanding characteristics. A range of additional measures to further increase the sensitivity are investigated with simulations.

We present a comprehensive noise model for an electromechanical resonator that is utilized as a magnetic field sensor. The cantilever-type sensor is coated with a magnetostrictive film that exhibits a change in elastic modulus E with a magnetic field and therefore detunes the resonator-the so-called delta-E effect. The noise model contains all relevant noise sources from the operational electronics, the wiring, and the sensor itself. Measurements show good agreement up to a certain excitation voltage, where an additional dominant noise source appears. It is identified as originating from the magnetic film. With the results of the model, the operational parameters of such sensors are discussed. The model predicts that the limit of detection at 10 Hz for the present sensors can be improved to 60 pT/(Hz)(1/2) if the magnetic noise is eliminated.

Thin-film magnetoelectric sensors reach a sensitivity in the picotesla range around the resonance frequency of the mechanical structure. Using magnetic frequency conversion, a magnetic low-frequency signal can be transferred to the sensor's resonance frequency. However, the required additional large carrier signal leaks to the sensor's output with a large amplitude, requiring a wide dynamic range of the sensor electronics. In this paper, it is shown that the unbalance of the magnetostriction curve is responsible for this leakage and that a suppression approach is devised. After theoretical analysis of the nonlinear magnetostriction characteristic, a carrier suppression is achieved through balancing by an altered signal excitation. A suppression of the carrier signal of about three orders of magnitude is measured. Thus, the requirements regarding analog-to-digital conversion can be reduced.

Thin-film magnetoelectric sensors are able to measure very low magnetic fields. As a consequence the hypothesis that magnetoelectric sensors could be used for biomagnetic measurements was often mentioned but never proven. In this contribution the first proof of this hypothesis will be given by the measurement of the (wellknown) R-wave of the human heart. This will be achieved by closing the gap between the sensor sensitivity and the signal level by averaging. In order to guarantee a fast convergence of the averaging process even in very noisy (realistic) measurement environments, different adaptive averaging techniques in the time-and frequency-domain are pointed out. The evaluation by synthetic measurements shows an improvement of the averaging process by up to 20 dB in terms of signal-to-noise ratio for an instationary measurement scenario in comparison to the conventional averaging after 750 average periods. Finally, measurements of the R-wave of a human heart are performed.

Thin-film magnetoelectric sensors, i.e., composites of magnetostrictive and piezoelectric materials, are able to measure very low magnetic flux densities in the picotesla range. In order to further improve the limit of detection it is of high importance to understand and quantify the relevant noise sources. In this paper, a common model for the deflection noise in vibrational structures is applied to the cantilever structure of resonant magnetoelectric sensors. By means of deflection and noise measurements the existence of thermal-mechanical noise even in sensor structures with a size in the centimeter range is proven. Based on these findings a noise equivalent circuit is suggested which allows not only the distinction between the impact of different sensor-intrinsic noise sources and also the involvement of the preamplifier noise. We found that the thermal-mechanical noise is the dominant noise source if direct signal detection is performed at the first bending resonance frequency of the sensor. However, this kind of noise is not the limiting influence when applying magnetic frequency-conversion techniques.

Composite magnetoelectrics implemented as thin film heterostructures are discussed in view of their applicability as highly sensitive magnetic field sensors. Here, either PZT or AlN served as piezoelectric component. The magnetostrictive phase consisted of layer systems based on FeCo or (Fe90Co10)(78)Si12B10. All functional layers were deposited with thicknesses of a few micrometers on Si cantilever structures with typical lateral dimensions of 25 mm by 2.2 mm. Magnetoelectric coefficients as large as 6900 V/cm Oe and a limit of detection as low as 1 pT/(Hz)(1/2) were measured. Currently, the best result demonstrates a detection limit of 500 fT/(Hz)(1/2) at 958 Hz frequency using a set of two sensors for external noise suppression. A frequency conversion technique is proposed to broaden the applicability of resonant magnetoelectric sensors to a wider frequency range. Finally, the achieved sensor performance is evaluated with regard to typical magnetic field amplitudes in medical applications.

Magnetoelectric thin film composites have demonstrated their potential to detect sub-pT magnetic fields if mechanical resonances (typically few hundred Hz to a few kHz) are utilized. At low frequencies (1–100 Hz), magnetic field-induced frequency conversion has enabled wideband measurements with resonance-enhanced sensitivities by using the nonlinear characteristics of the magnetostriction curve. Nevertheless, the modulation with a magnetic field with a frequency close to the mechanical resonance results in a number of drawbacks, which are, e.g., size and energy consumption of the sensor as well as potential crosstalk in sensor arrays. In this work, we demonstrate the feasibility of an electric frequency conversion of a magnetoelectric sensor which would overcome the drawbacks of magnetic frequency conversion. This magnetoelectric sensor consists of three functional layers: an exchange biased magnetostrictive multilayer showing a high piezomagnetic coefficient without applying a magnetic bias field, a non-linear piezoelectric actuation layer and a linear piezoelectric sensing layer. In this approach, the low frequency magnetic signal is shifted into the mechanical resonance of the sensor, while the electric modulation frequency is chosen to be either the difference or the sum of the resonance and the signal frequency. Using this electric frequency conversion, a limit of detection in the low nT/Hz1/2 range was shown for signals of low frequency.

Thin-film magnetoelectric (ME) sensors offer a promising potential to measure biomagnetic signals in the near future. Unfortunately, this sensor type shows usually a large cross-sensitivity to all kinds of mechanic distortion due to the resonant structure. In order to overcome this problem several sensor designs have been proposed. Beside these approaches adaptive noise cancellation techniques can be used to reduce the noise coupling while keeping the sensor setup simple. In this contribution reference sensors, realized as piezoelectric cantilevers, are presented and compared to microphones by means of their feasibility to improve the signal-to-noise ratio (SNR) using adaptive cancellation approaches. If a loudspeaker is used as noise source, no crucial differences are measured. But if a vibrator is used as noise source to generate structure-borne noise, the piezoelectric (PE) cantilevers are superior. As the difference of the resonance frequencies between ME and PE sensor is decreased the SNR improvement increases at low excitation levels. In total an SNR improvement over 30 dB can be achieved.

Thin-film magnetoelectric sensors, i.e., composites of magnetostrictive and piezoelectric materials, are able to measure very low magnetic fields. As a consequence, an application of such sensors could be, e.g., the measurement of biomagnetic fields in the near future. To measure these signals, typically characterized by low-frequency components, techniques, such as the delta-E effect, are utilized. The limit of detection (LoD) of such sensor systems did not reach the required level until now. In order to improve this, an adaptive readout scheme is proposed for sensor systems based on the delta-E effect. The basis is a simultaneous measurement with a single sensor at different frequency ranges close to the resonance frequencies. The signals are combined optimally in regard to their signal-to-noise ratio. Two combination approaches are presented and evaluated. An improvement up to 6 dB in terms of LoD is achieved. Due to an adaption of the weighting coefficients with time, the proposed method can be interpreted as a noise reduction technique, which increases the usability of such sensors in realistic measurement scenarios.

We present an analytical and experimental study on low-noise piezoelectric thin film resonators that utilize the delta-E effect of a magnetostrictive layer to measure magnetic fields at low frequencies. Calculations from a physical model of the electromechanical resonator enable electrode designs to efficiently operate in the first and second transversal bending modes. As predicted by our calculations, the adapted electrode design improves the sensitivity by a factor of 6 and reduces the dynamic range of the sensor output by 16 dB, which significantly eases the requirements on readout electronics. Magnetic measurements show a bandwidth of 100 Hz at a noise level of about 100 pTHz−0.5.

Monitoring driver's intentions beforehand is an ambitious aim, which will bring a huge impact on the society by preventing traffic accidents. Hence, in this preliminary study we recorded high resolution electroencephalography (EEG) from 5 subjects while driving a car under real conditions along with an accelerometer which detects the onset of steering. Two sensor-level analyses, sample entropy and time-frequency analysis, have been implemented to observe the dynamics before the onset of steering. Thus, in order to classify the steering direction we applied a machine learning algorithm consisting of: dimensionality reduction and classification using principal-component-analysis (PCA) and support-vector-machine (SVM), respectively. The results showed an increase of the sample entropy and the estimated power values in the theta and alpha frequency bands, 100 ms before the onset of steering. The detection of steering direction depicted that sample entropy gives a higher classification accuracy (73.5% ±6.8) as compared to that of using the estimated power for theta and alpha frequency bands (62.6% ±5.6).

Sensors based on the magnetoelectric (ME) effect have the potential to be genuine alternatives for measuring bio-magnetic signals. Unfortunately, the sensor structure usually inhibits the problem that several non-magnetic types of noise couple mechanically into the sensor: in this contribution, we will focus on undesired acoustic coupling. Therefore, an adaptive cancellation approach based on a computationally efficient gradient estimation algorithm with a pseudo-optimally control scheme is proposed. The approach is using a microphone as a noise reference sensor and is implemented in real time. An evaluation in terms of measurements is performed inside a magnetically shielded chamber. For a particular scenario, which is characterized by double excitation, an algorithm with binary control-scheme improves the signal-to-noise ratio (SNR) only by around 4dB. If the proposed control scheme is used instead, an improvement of the SNR of around 13dB is achieved.

For magnetic field measurements with magnetoelectric sensors sensitivities for AC magnetic fields in the femto-Tesla range are desired. Important factors for the necessary signal/noise ratio are extrinsic noise sources such as vibrations and acoustics. In this paper a concept to reduce the noise contributions from these sources is suggested and investigated. It is shown that by connecting two magnetoelectric sensors with pico-Tesla sensitivity in a tuning fork structure the noise level caused by acoustic and vibrational coupling can be strongly reduced. (C) 2015 Elsevier B.V. All rights reserved.

Recently, magnetoelectric sensors have gained importance in medical applications. Among several sensor types microelectromechanical magnetic field sensors are a promising option since they can be manufactured in a very small fashion and can operate at room temperature. In this contribution we present a new readout scheme for such sensors that simultaneously uses amplitude and phase information of a multitude of mechanical modes of a single sensor. The sensors are based on the delta-E effect. All extracted information is optimally combined in order to get a sensor output signal with maximum signal-to-noise ratio (SNR). First measurements have shown very promising results.