betaFeedback

WorksPhysics and Astronomy

Development of an integrated device based on the gain/phase detector and Arduino platform for measuring magnetoelastic resonance

Abstract

This study presents the development of an integrated device for measuring magnetoelastic resonance, utilizing a gain/phase detector and the Arduino platform. The device stands out for its simplicity, low cost, and efficiency. It employs the Arduino's ATmega328P microcontroller, coupled with the AD8302, and an intuitive graphical interface developed in Python, which facilitates the measurement of magnetoelastic signals in sensors. The device's validation is carried out through measurements of mass deposits on magnetoelastic sensors, confirming its accuracy and functionality. This advancement offers a practical and portable solution for detecting magnetoelastic resonances, promoting the use of these technologies in field environments and in diverse scientific applications. The integration of components into a compact, robust, and low-cost electronic circuit, along with an easy-to-operate graphical interface, significantly simplifies the complexity of the components, making the equipment complete and readily operational.

Introduction

Magnetoelastic materials (ME) stand out for their notable potential in the development of sensors that play a crucial role in the scenario of scientific research related to the detection of physical, chemical and biological agents, as evidenced by recent studies [1]. Sensors manufactured with these materials operate based on the phenomenon of magnetostriction, in which the magnetic response of the material is directly influenced by mechanical deformations. The ability of these materials to convert changes in shape, resulting from applied voltages, into measurable variations in their magnetic state, gives these sensors high sensitivity and precision. The main advantages of which are: i) the possibility of wireless detection, ii) high sensitivity [2, 3, 4], iii) chemical and physical resistance [5, 6]. Furthermore, it is a versatile and easily viable technique for its production and detection, dispensing the need for electrical contacts and lithographic processes. Added to this, the remote query nature of the ME sensor platform can offer a major advantage when a direct probe or electrical contact with the sensor is not a viable alternative.

Given the growing importance of detection platforms based on ME materials, the need to develop instrumentation for the detection of ME signals becomes evident. Laboratory devices such as vector network analyzers (VNA) [3], real-time spectrum analyzers (RTSA) [7], impedance analyzers [8], Lock-in amplifier application circuits [9] and oscilloscopes [10] are some of the instruments that can be used as detectors of these signals. However, the applicability of these instruments in field environments is limited due to their weight, high cost and complexity. Therefore, the search for more practical and accessible solutions for implementing this technology is essential, aiming for its effective implementation in real-world conditions. Several studies have been dedicated to this issue [11, 12, 13, 14]. Recently Sang et al. [15] proposed a portable microcontroller-based system, integrated through System on Chip (SoC) architecture, to sample resonance frequency changes in an ME sensor used in the detection of human serum albumin (HSA). Shen et al. [16] and Zeng et al. [14] proposed time-domain instrumentation to determine the resonance frequency in ME sensors. Yang et al. [17] used a portable system for measurements in aquatic environments.

However, these devices feature relatively complex electronic circuits, requiring significant infrastructure for manufacturing, making them impractical in scenarios where cost efficiency and ease of production are priorities. On the other hand, conventional laboratory equipment can require a complex

Implementation of the electronic circuit and graphical user interface (GUI)

The methodology adopted to perform resonance frequency measurements is a frequency sweep. Arduino's ATmega328P microcontroller performs integral control of the electronic system. The AD9833 generates a sinusoidal signal to sweep the excitation coil, which initially passes through a preamp followed by a power amplifier. The excitation circuit couples a DC signal (induction) to the AC sinusoidal signal (modulation), through a power amplifier. The DC signal is necessary to generate a magnetic induction field, so that the ME sensor oscillates with greater amplitude in its anisotropy field. A pickup coil is monitored by a phase and gain detector, which receives a reference signal from the signal generator. Thus, the values associated with the magnitude and phase of the signal are obtained by the pickup coil. When inserting the sensor inside, it enters resonance when the frequency of the excitation signal corresponds to its magnetoelastic resonance frequency, identified as a descending amplitude peak in the signal magnitude spectrum. The entire circuit is powered by a symmetrical 9 V supply. The Arduino platform exchanges data and is powered with +5+5 V through the computer's USB port. An electronic schematic of the main components of the developed device are detailed in Figure 1.

Image not extracted

Figure 1. Electronic diagram with the main components of the developed device. (1) The ATmega328P microcontroller controls the signal from the AD9833 (3), reads the signal from the AD8302 (9) and displays the scan parameters on the LCD display (2). Interaction is enhanced through the Graphical User Interface (GUI), allowing adjustments to scan intervals and DC and AC field strengths. Graph (13) shows the linear response to the signal generated by the AD9833. In (10) there is the signal output generated by the AD8302 as a function of the logarithm of the relationship between the signals, VINPAV_{\mathrm{INPA}} and VINPBV_{\mathrm{INPB}}, in (11), the signal output related to the phase difference (θINPA−θINPB\theta_{\mathrm{INPA}} - \theta_{\mathrm{INPB}}) and, in (12), the coupled signal (DC + AC) delivered to the excitation coil.

ATmega328P Microcontroller: the ATmega328P (1) present on the Arduino board is an 8-bit microcontroller with RISC architecture and a clock speed of up to 20 MHz. With integrated peripherals, such as 10-bit resolution ADC and UART serial communication, it offers robust features for control and measurement. Its accessible programming via the Arduino platform makes it ideal for instrumentation projects, being able to easily interact with external sensors and devices. The Arduino is used to establish communication between the excitation (3, 4, 5 and 6) and detection (7 and 9) circuits, as well as carrying out serial communication with the microcomputer. The ATmega328P also receives a 1.8 V signal from the AD8302 (15) for ADC reference, resulting in a resolution of 1.76 mV. The microcontroller is programmed in C++ language through the Arduino IDE, to communicate with the peripherals and the developed graphical interface.

AD9833 Programmable Waveform Generator: The AD9833 (2) is a low power programmable waveform generator (12.65 mW power consumption at 3 V), capable of producing sine, triangle and square wave outputs. The output frequency and phase are programmable. The frequency registers have a width of 28 bits, allowing a resolution of 0.1 Hz to be achieved with a clock frequency of 25 MHz, and a frequency range of 0 to 12.5 MHz [18]. The device is programmed to generate sine waves, which are produced with an amplitude of 0.6 V and undergo subsequent amplification processes.

Pre and power amplifier: the signal from the AD9833 does not have the ability to excite the coil to stimulate sensor oscillation. Therefore, it is necessary to amplify this signal. A preamplifier (4), based on the LM358 operational amplifier, is used to increase the amplitude of the signal generated by the AD9833 from 0.6V to 1.2V. The signal then goes to a second amplification stage, using the OPA544 operational power amplifier (5). At this stage, it becomes possible to adjust the amplitude of the AC signal from 1.2 V to 8.0 V, with considerable current gain. A DC offset signal (necessary to generate the induction magnetic field) is added to the AC signal, allowing adjustments in the range of -5 V to +5 V. The resulting new AC signal is then filtered and sent to the detection system for reference. The AD8302 (only the AC component follows for reference), which is also injected into the excitation coil, generates the induction (DC component) and modulation (AC component) magnetic fields, responsible for exciting the sensor.

Excitation and pickup coils: after amplification of the signal, it is applied to the excitation coil (6) (600 turns, 23 mm internal diameter and 50 mm long, with 0.40 mm thick copper wire and resistance of 10.7 Ω), which produces the magnetic field (HDC+HACH_{\mathrm{DC}} + H_{\mathrm{AC}}) necessary to generate the oscillations in the ME sensor. A double-layer pickup coil (7) (200 turns, 4 mm inner diameter and 22 mm long, with 0.18 mm thick copper wire and 4.5 Ω\Omega resistance) is allocated concentric to the excitation coil, to which is monitored by the detection circuit described below. A calibration was performed to obtain the DC induction magnetic field from the signals read by the ADS on port A1.

Phase and gain detector (GPD): The GPD (9) integrate the device's detection circuit, based on the AD8302 chip from Analog Devices, which measures the magnitude ratio (gain or loss) and the phase difference between two signals [19]. The AD8302 circuit (9) operates as a single-chip vector signal analyzer, where two monolithically integrated identical logarithmic amplifiers play main roles. Such amplifiers perform a logarithmic compression of the input signals, VINPAV_{\mathrm{INPA}} and VINPBV_{\mathrm{INPB}}, which can vary from -60 dBm to 0 dBm, generating two voltage signals at the output, VMAGV_{\mathrm{MAG}} and VPHSV_{\mathrm{PHS}}, with a linear range of 0 to 1.8 V. The component also incorporates a subtractor to determine the difference between the input signals. This makes it possible to accurately measure the amplitude and phase of the signal. The operating range extends to 2.7 GHz, with magnitude ratio measurement in a range of ±30\pm30 dB, at a sensitivity of VSLP=30V_{\mathrm{SLP}} = 30 mV/dB (600 mV/decade), and the phase difference measurement range between the input signals is from 0∘0^\circ to 180∘180^\circ, at a sensitivity of Vϕ=10V_{\phi} = 10 mV/∘^\circ, which further expands the usefulness of this device in various applications that require detailed analysis of radiofrequency (RF) signals. With the AD8302 operating in “Measurement Mode”, the expression for the VPHSV_{\mathrm{PHS}} output phase and the VMAGV_{\mathrm{MAG}} magnitude gain are given by equations (1) and (2), respectively [19, 20]:

VMAG=VSLPlog⁡(VINPAVINPB)+VCP,(1)V_{\mathrm{MAG}} = V_{\mathrm{SLP}} \log\left(\frac{V_{\mathrm{INPA}}}{V_{\mathrm{INPB}}}\right) + V_{\mathrm{CP}}, \tag*{(1)}
VPHS=±Vϕ(ϕ(∘)−90∘)+VCP,(2)V_{\mathrm{PHS}} = \pm V_{\phi}\left(\phi(^\circ) - 90^\circ\right) + V_{\mathrm{CP}}, \tag*{(2)}

where ϕ(∘)=ϕINPA−ϕINPB\phi(^\circ) = \phi_{\mathrm{INPA}} - \phi_{\mathrm{INPB}} is the phase difference between the input signals. For VMAGV_{\mathrm{MAG}}, the midpoint is VCP=900V_{\mathrm{CP}} = 900 mV for 0 dB gain, and a range of −30-30 dB to +30+30 dB covers the full swing from 0 V to 1.8 V, as shown in the graph (10) in Figure 1. For the VPHSV_{\mathrm{PHS}} phase function, the center point is VCP=900V_{\mathrm{CP}} = 900 mV for 90∘90^\circ, and a range from 0∘0^\circ to 180∘180^\circ covers the full swing from 1.8 V to 0 V, as shown in Figure 1 (11). Equations (1) and (2) can be rewritten to express the magnitude Mag (dB) and the phase difference ϕ(∘)\phi(^\circ) [21, 22]:

Mag(db)=log⁡(VINPAVINPB)=VMAG(mV)−900mV30mV.(3)Mag(db) = \log\left(\frac{V_{\mathrm{INPA}}}{V_{\mathrm{INPB}}}\right) = \frac{V_{\mathrm{MAG}}(mV) - 900mV}{30mV}. \tag*{(3)}
ϕ(∘)=ϕINPA−ϕINPB=±(VPHS(mV)−900 mV10 mV/∘+90∘).(4)\phi(^{\circ}) = \phi_{INPA} - \phi_{INPB} = \pm\left(\frac{V_{PHS}(mV) - 900\ mV}{10\ mV/^{\circ}} + 90^{\circ}\right). \tag*{(4)}

Symmetrical power supply: To provide symmetrical power supply to the amplification circuit, a symmetrical power supply (1) was designed based on LM7809 and LM7909 voltage regulators. These regulators generate a fixed voltage value of ±9\pm9 V, ensuring a complete wave cycle in the sinusoidal signal. The electronic circuit design, integrating all circuit components, is shown in Figure 2.

Image not extracted

Figure 2. Electronic circuit of the device. In (1), ATmega328P microcontroller and (2) LCD display, (3, 4 and 5) AC and DC signal generation circuit. At this stage, the microcontroller's analog terminations A0A_0 and A1A_1 read the AC and DC fields, respectively. Excitation (6) and pickup (7) coils, coil signal reading circuit (9). At this stage, the microcontroller terminations A2A_2 and A3A_3 read the phase and magnitude of the signal read from the pickup coil. Finally, in (14) power supply.

Device assembly

The final device was built by placing the entire electronic circuit in plastic cases. The numbers in Figure 2 correspond to the component numbers described in Figure 1. The LCD display (2) was fixed in one of the cases, measuring 12×8×412 \times8 \times4 centimeters, together with the excitation and detection circuit, which has its reference, excitation and capture signal outputs connected through SMA terminals. In a second case, measuring 7×4×37 \times4 \times3 centimeters, the excitation coil (6) and the pickup coil (7), were fixed, both also equipped with SMA terminals. Communication between the electronic control unit and the coils is established using coaxial cables with a characteristic impedance of 50 Ω50~\Omega. The symmetrical source circuit (14) was placed in a third plastic case, measuring 9×4×49 \times4 \times4 centimeters, which communicates with the electronic control unit for power supply via banana-terminated cables. This organization in cases provides a resistant and robust structure for the device, facilitating assembly, maintenance and adaptation as necessary.

Developed graphical user interface (GUI)

The program, developed in the open-source programming language Python, makes the equipment intuitive and easy to operate. For its production, the Tkinter libraries were used for graphic creation, Matplotlib for displaying graphics and Pyautogui for capturing screens. Additionally, libraries such as Serial, Numpy and Scipy are imported for data manipulation and serial communication. The program includes fields for entering parameters, such as serial port address, start and end frequency, sweep step, amplitude and start and end phase. During the scan, data is read from the serial port, processed and displayed in real time on graphs (amplitude and phase) to accompany the scan, where it is also possible to monitor values associated with the DC magnetic field, AC signal amplitude and quality factor of the Q resonance, in addition to saving all data in a .txt file and having the option of exporting images of the graphs in .png format. The image of the main screen of the developed GUI is shown in Figure 3. A link with the executable version of the graphical interface, together with the Arduino programming code, is available at https://github.com/wphysics/WaveMagScan.

Image not extracted

Figure 3. Overview of the graphical user interface (GUI) developed for communication with the developed device, through which you can control the parameters associated with the frequency sweep, monitor the sweep through the magnitude phase difference graph, as well as view the values induction fields (DC), modulation (AC) and resonance quality factor. A file in txt format with all these parameters is automatically saved with each scan.

ME sensor preparation and measurement

Three ME sensors were produced from ribbons of size 11 mm×2 mm×28 μm11\ \mathrm{mm}\times2\ \mathrm{mm}\times28\ \mu\mathrm{m} of METGLAS® 2826MB3 (Fe40Ni38Mo4B18\mathrm{Fe}_{40}\mathrm{Ni}_{38}\mathrm{Mo}_{4}\mathrm{B}_{18}) alloy. The sensors were cleaned in ultrasonic baths, first in acetone and then in ethanol, for 10 minutes in each step. Subsequently, a Platinum (Pt) film was deposited on the polished side of five sensors, with thicknesses of 50, 100, 150, 200 and 250 nm, respectively. The resonance frequencies of each ME sensor were measured using the proposed equipment, with a frequency sweep step of 10 Hz. To measure the magnetoelastic resonance frequency, the sensor is inserted inside the pickup coil. Then, through the user interface, the measurements are initialized. All measurements were performed at room temperature (25 °C). For comparison purposes, resonance frequency measurements were performed using a vector network analyzer (R&S®ZNLE18, Rohde & Schwarz, Munich, Germany), operated in S11_{11} mode connected to an internal solenoid coil (the detection coil), which generates the AC excitation signal with a step of 5 Hz, to perform a frequency sweep and monitor the reflected signal. The DC magnetic induction field was applied through the external coil (the excitation coil) using a Keysight U8031A DC source.

ME resonance frequency detection principle and measurements

Due to a strong coupling between mechanical and magnetic energies, ME materials can change their dimensions in the presence of a magnetic field. The reflections of which can be studied by evaluating the modulus of elasticity EE of the ME material, which will be subject to the “delta EE” effect, where the modulus of elasticity is a function of the applied magnetic field HH, E=E(H)E = E(H), given by equation (5) [16, 23] [24]:

E(H,σ)=(1EH+9λs2H2μ0MsHAσ3)−1(5)E(H,\sigma) = \left(\frac{1}{E_H} + \frac{9\lambda_s^2 H^2}{\mu_0 M_s H_{A\sigma}^3}\right)^{-1} \tag*{(5)}

where EHE_H is the modulus of elasticity without the effect of the magnetic field (H=0H = 0), μ0\mu_0 is the magnetic permeability of the vacuum, λs\lambda_s is the magnetostriction constant, MsM_s is the saturation magnetization and HAσH_{A\sigma} is the magnetoelastic anisotropy field, where the sensor reaches the highest oscillation amplitudes, which depends on the voltage σ\sigma applied to the material.

A plane wave type disturbance is performed on the ME sensor through the application of AC (modulation) and DC (induction) magnetic fields, H=HDC+HACH = H_{\mathrm{DC}} + H_{\mathrm{AC}}. By ensuring that the magnetic fields are uniform along the sample (x axis), a one-dimensional ME wave will propagate through the material, which can be described by equation (6):

∂2ux∂t2=Eρr(1−ν2)∂2ux∂x2(6)\frac{\partial^2 u_x}{\partial t^2} = \frac{E}{\rho_r(1-\nu^2)}\frac{\partial^2 u_x}{\partial x^2} \tag*{(6)}

whose solution is of the type [25]:

ux(t)=u0cos⁡(nπxL)e−iωt(7)u_x(t) = u_0 \cos\left(\frac{n\pi x}{L}\right)e^{-i\omega t} \tag*{(7)}

where ω0\omega_0 is a constant and ω=2πf\omega= 2\pi f is the angular frequency of oscillation of the applied AC field, nn is the resonance mode (evaluated at n=1n = 1), LL is the length of the density sensor. From equation (6) and (7), the ME resonance frequency frf_r can be located by equation (8) [1] [25]:

fr=n2LEρ(1−ν2)(8)f_r = \frac{n}{2L}\sqrt{\frac{E}{\rho(1-\nu^2)}} \tag*{(8)}

where ρ\rho is the density and ν\nu is the Poisson's ratio of the ME sensor. Considering a layer of mass sufficiently thin in relation to the wavelength of the ME resonator, rigidly deposited and uniformly distributed over its surface, which oscillates synchronously with the resonator, so that the mass layer can be treated as part of the resonator. Under these conditions, the average density of the resonator charged with mass ρc\rho_c will be:

ρc=ρr+ρmhd(9)\rho_c = \rho_r + \frac{\rho_m h}{d} \tag*{(9)}

where ρm\rho_m and hh are the density and thickness of the deposited mass layer and dd is the thickness of the ME resonator. Solving equation (6) for (ρr=ρc)(\rho_r = \rho_c), a relationship between the perturbed ω\omega and the free ω0\omega_0 resonance frequencies by the deposited mass is obtained, given by:

(ωω0)2=(1+ρmhρrd)−1(10)\left(\frac{\omega}{\omega_0}\right)^2 = \left(1 + \frac{\rho_m h}{\rho_r d}\right)^{-1} \tag*{(10)}

For a thin layer of dough (ρmh≪ρrd\rho_m h \ll\rho_r d), equation (10) can be approximated by a series expansion in a linear relationship between frequency change Δf\Delta f and the property ρmh\rho_m h of the dough layer [26]:

Δfmass=−ρmh2ρrdf0(11)\Delta f_{mass} = -\frac{\rho_m h}{2\rho_r d}f_0 \tag*{(11)}

where Δfmass\Delta f_{mass} is the displacement of the natural frequency f0f_0 of the ME resonator, resulting from the deposition of a thin film on its surface. Assuming that the increase in mass Δm=ρmhA\Delta m = \rho_m h A occurs over the entire surface area AA of the ME resonator, equation (11) takes the form [1]:

Δf=−Δm2m0f0(12)\Delta f = -\frac{\Delta m}{2m_0}f_0 \tag*{(12)}

where m0m_0 is the mass of the bare ME resonator. Finally, the mass sensitivity is defined in equation (13).

Higher resonance frequencies, which can be achieved with sensors of shorter length LL, or sensors of lower mass m0m_0 result in higher sensitivity values.

sm=−ΔfΔm=f02m0.(13)s_{\mathrm{m}}=-\frac{\Delta f}{\Delta m}=\frac{f_0}{2m_0}. \tag*{(13)}

Device developed

Figure 4 shows the completed device. It integrates all the parts necessary for the detection of magnetoelastic signals in a single unit, such as excitation coils, detection, signal reading and control center, symmetric power supply, RF voltage gain control (modulation) and DC signal (offset). The ability to program scan settings and control the device through the GUI is an advantage that adds flexibility to the measurement and analysis process, enabling quick and easy control and recording of data. Relevant parameters in the scanning process are managed directly on its main GUI screen, such as the start frequency, end frequency and sweep step, allowing detection of ME signals with high resolution down to 0.1 Hz over a wide range of frequencies (100 Hz to 25 MHz). The induction and modulation field, resonance frequency and resonance quality factor Q are displayed on the screen, simplifying control and data recording and offering greater flexibility to customize measurements to the specific needs of each application. Table 1 summarizes the range and resolution achieved with the proposed device for ME resonance measurements.

Image not extracted

Figure 4. Overview of the developed prototype device. (a) symmetrical 9 V power supply, (b) electronic control unit with LCD display, (c) microcomputer, auxiliary in the detection process, (d) detection and (e) excitation coils, internal and concentric to the coil of excitation (d), where the ME sensor is positioned for reading. (f) Image of an oscilloscope screen of the output signals in blue (excitation), on a scale of 2 V per division, and input signal in yellow (detection), with a scale of 0.1 V per division.

ParametersRangeResolution
Frequency (excitation)0 a 12.5 MHz0.1 Hz
HDCH_{\mathrm{DC}} (excitation)-30 a 30 Oenot measured
AC amplitude (excitation)0.2 a 3.0 Vnot measured
Phase difference (detection)0 a 180°10 mV/°
Magnitude (dB) (detection)−30 a 30 dB30 mV/dB

Table 1. Range and resolution of measurements with the developed device.

The excitation, capture and reference signals were analyzed, which can be visualized in Figure 5. Figure 5a shows the sinusoidal signal in blue, coming from the excitation system, with a peak-to-peak amplitude of 3.08V, obtained after going through the amplification stages. In yellow, is the reference signal, which passes through a voltage divider to have an amplitude equivalent to the signal captured by the pickup coil, presenting a peak-to-peak amplitude of 0.460V. Both signals in Figure 5a are in phase. In Figure 5b, the reference (yellow) and capture (blue) signals are shown. The capture signal is 90∘90^\circ ahead of the reference signal, as can also be seen in Figure 4f for the excitation signal. This phase change is the result of Lenz's law, as the flux induced in the pickup coil by the excitation coil has the opposite direction, causing the signal to phase change [27]. This signal phase shift is measured by the detection system (as shown in Figure 7b).

Image not extracted

Figure 5. Visualization of the main signals throughout the excitation circuit. (a) Excitation (blue) and reference (yellow) signals in phase. (b) Capture (blue) and reference (yellow) signals with a phase difference of 90∘90^\circ, a change attributed to Lenz's Law.

Image not extracted

Figure 7. (a)-(b) Results of measuring the magnetoelastic resonance frequency of the ME sensor measuring 11 x 2 x 0.03 mm, carried out with the proposed equipment and (c)-(d) the VNA in an anisotropy field HAσ=13.5H_{A\sigma} = 13.5 Oe. Both configurations show differences in the curves, due to the different experimental approaches. However, accuracy in determining the resonance frequency is obtained, highlighting the reliability of the proposed equipment.

The static magnetic field HDCH_{DC} along the x-direction generated by the DC bias coil is determined from the Biot-Savart law:

HDC(x)=μ0inDC2[L2−x(L2−x)2+R2+L2+x(L2+x)2+R2],(14)H_{DC}(x) = \frac{\mu_0 i n_{DC}}{2}\left[\frac{\frac{L}{2-x}}{\sqrt{\left(\frac{L}{2-x}\right)^2+R^2}}+\frac{\frac{L}{2+x}}{\sqrt{\left(\frac{L}{2+x}\right)^2+R^2}}\right], \tag*{(14)}

where ii is the electric current, nDCn_{DC} is the number of turns, LL and RR are the length and the radius of the DC bias coil, respectively. Figure 6a shows the DC polarization coil. Figure 6b represents the magnetic field lines generated by it in the x-y plane. The graph in Figure 6c shows the fit of equation (14) to experimental data for the HDCH_{DC} magnetic field along the x axis. In the region shaded in blue in Figure 6c, the magnetic field remains constant and homogeneous throughout the ME sensor, ensuring ideal excitation conditions.

Image not extracted

Figure 6. (a) Image of the DC bias coil. At its center is the pick-up coil. (b) Representation of the magnetic field lines generated by the DC bias coil in the x-y plane. (c) Blue line is the fit the experimental data to equation (14). The region comprised by the outermost blue shadow (Figure 6c) is related to the length of the DC polarization coil, and the inner region is related to the length of the internal coil.

Measurements obtained with the proposed device

The results of measuring the magnetoelastic resonance frequency of the ME sensor are shown in Figure 7, carried out with the proposed equipment in this work (Figure 7a-Figure 7b) and with the VNA (Figure 7c-Figure 7d). Both measurements were performed in the anisotropy field HAσ=13.5H_{A\sigma} = 13.5 Oe. In the case of different experimental configurations, when measuring with the proposed equipment, excitation occurs by the external coil and capture by the internal coil, while the VNA uses the internal coil for both excitation and capture, and the external coil is used to generate the HDCH_{DC} induction field. Therefore, although the curves and quantities analyzed present differences, the magnetoelastic resonance frequency is precisely determined in both assemblies, demonstrating the accuracy of the proposed equipment in these measurements. The resonance frequency measured by the proposed equipment is obtained at the minimum magnitude value (Figure 7a). The resonance frequency is located at the center of the phase difference curve.

The graphs in Figure 8 show the results of measurements carried out with the proposed equipment of (Figure 8a) magnitude and (Figure 8b) phase difference measured by a sensor measured in air and immersed in water. It is verified that there is a shift in the resonance frequency as well as a drop in the oscillation amplitude measured in the sensor immersed in water when compared to the measurements carried out in air. This is due to the viscous friction that the liquid generates on the sensor surfaces, which also reflects a decrease in the resonance quality factor QQ [28-29], where:

Image not extracted

Figure 8. (a) Magnitude and (b) phase difference measured from a sensor in air and immersed in water. (c) Variation of the resonance frequency as a function of the applied magnetic field HH. Such a change in the resonance frequency is attributed to the "delta E" effect, where the Young's modulus is a function of HH, directly impacting the elastic properties of the material and, consequently, the system's resonance frequency.

Q=frΔf→QAir=552.5QWater=144.5(15)Q = \frac{f_r}{\Delta f} \quad\rightarrow\quad\begin{aligned} Q_{Air} &= 552.5 \\ Q_{Water} &= 144.5 \end{aligned} \tag*{(15)}

where Δf\Delta f is the width at half height on the magnitude curve. The graph in Figure 8c represents the variation in the resonance frequency as a function of the applied magnetic field HH. The variation observed in the resonance frequency is attributed to the "delta E" effect, indicating that the Young's modulus of the ME sensor is influenced by the magnetic field, a characteristic effect of magnetoelastic materials. From the analysis of the curve in Figure 8c, we arrive at the value of the anisotropy field, obtained in the valley, whose value was found to be HAσ=13.5H_{A\sigma} = 13.5 Oe, used in this work [30].

The shift in the ME resonance frequency as a function of the platinum (Pt) deposit on the ME sensor was analyzed using the proposed device. The graph in Figure 9a shows the presence of the shift in the ME resonance frequency for different thicknesses of Pt deposited on the sensor, always moving to the left. A small downward shift at the base of the curve signals (50 - 250 nm) in relation to the signal without deposition (0 nm) is due to a possible influence of the deposited film on the magnetic susceptibility of the medium, which is measured by the log⁡(VINPAVINPB)\log\left(\frac{V_{INPA}}{V_{INPB}}\right) expressed in equation (1). A small drop in signal amplitude for depositions (50 - 250 nm) in relation to the signal without deposition (0 nm) is also observed, a possible contribution from the tension formed between the substrate and the deposited film, which may lead to this decrease in amplitude.

Image not extracted

Figure 9. (a) Shift of the ME resonance frequency as a function of the thickness of the deposited film. (b) Absolute shift of the ME resonance frequency for different masses associated with films of different thicknesses deposited.

The graph in Figure 9b shows the absolute displacement of the ME resonance frequency for different masses associated with films of different thicknesses deposited, considering the sensor area A=0.00005 m2A = 0.00005\ \mathrm{m}^{2} and the density of platinum ρPt=21500 Kgm3\rho_{Pt} = 21500\ \frac{\mathrm{Kg}}{\mathrm{m}^{3}}. In this way, we can arrive at the sensitivity factor E described by equation (13), whose value obtained was sm=20.13 Hzμgs_m = 20.13\ \frac{\mathrm{Hz}}{\mu\mathrm{g}}. The sensitivity sms_m is directly related to the mass of the bare sensor, as described in equation (13), that is, the smaller the sensor, the more sensitive it is. Close values for sms_m, considering the size of the sensor used. Saiz et al. [1] obtained values of sm=47.4 Hzμgs_m = 47.4\ \frac{\mathrm{Hz}}{\mu\mathrm{g}} for a sensor using the same Metglas 2826 MB alloy with dimension 1 mm×10 mm×28 μm1\ \mathrm{mm} \times10\ \mathrm{mm} \times28\ \mu\mathrm{m} and sm=12.3 Hzμgs_m = 12.3\ \frac{\mathrm{Hz}}{\mu\mathrm{g}} for a sensor dimension 1.5 mm×15 mm×28 μm1.5\ \mathrm{mm} \times15\ \mathrm{mm} \times28\ \mu\mathrm{m}, for different deposited gold (Au) thicknesses. Ersöz et al. [31] using the same Metglas 2826MB alloy with dimensions of approximately 55 mm×6.5 mm×30 μm55\ \mathrm{mm} \times6.5\ \mathrm{mm} \times30\ \mu\mathrm{m}, found lower sensitivity values, around sm=1 Hzμgs_m = 1\ \frac{\mathrm{Hz}}{\mu\mathrm{g}}.

Patent

This device and the GUI software are registered with the Brazilian National Institute of Intellectual Property (INPI, Ministry of Economy, Brazil) under registration numbers BR102024008915-4 and BR512024000754-7, respectively.

Conclusions

The study presented the development of an integrated device based on a gain/phase detector and the Arduino platform for measuring magnetoelastic resonance, standing out for its simplicity, low cost, and efficiency. Utilizing the ATmega328P microcontroller of the Arduino coupled with the AD8302 and an intuitive graphical interface developed in Python, the device facilitates the measurement of magnetoelastic signals in sensors, making the technology accessible to various fields of research. The validation of the device was carried out through measurements of mass deposits on ME sensors, confirming its accuracy and functionality. This advancement provides a practical and portable solution for detecting magnetoelastic resonances, promoting the use of these technologies in field environments and other scientific applications.

Acknowledgements

This research is supported by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Financiadora de Estudos e Projetos (FINEP), Fundação de Amparo à Pesquisa do Estado de Minas Gerais (FAPEMIG) - Rede de Pesquisa em Materiais 2D and Rede de Nanomagnetismo, and INCT of Spintronics and Advanced Magnetic Nanostructures (INCT-SpinNanoMag), No. CNPq 406836/2022-1.

References

  1. [1]P. G. Saiz, R. Fernández de Luis, A. Lasheras, M.I. Arriortua, A.C. Lopes, Magnetoelastic resonance sensors: Principles, applications, and perspectives, ACS Sens. 7 (2022) 1248–1268. https://doi.org/10.1021/acssensors.2c00032.
  2. [2]S. Sang, X. Guo, J. Wang, H. Li, X. Ma, Real-time and label-free detection of VKORC1 genes based on a magnetoelastic biosensor for warfarin therapy, J. Mater. Chem. B Mater. Biol. Med. 8 (2020) 6271–6276. https://doi.org/10.1039/d0tb00354a.
  3. [3]R. Liu, X. Guo, J. Wang, J. Guo, Y. Zhang, W. Zhang, S. Sang, High sensitivity detection of human serum albumin using a novel magnetoelastic immunosensor, J. Mater. Sci. 54 (2019) 9679–9688. https://doi.org/10.1007/s10853-019-03554-0.
  4. [4]J. Wang, X. Guo, R. Liu, J. Guo, Y. Zhang, W. Zhang, S. Sang, Detection of carcinoembryonic antigen using a magnetoelastic nano-biosensor amplified with DNA-templated silver nanoclusters, Nanotechnology 31 (2020) 015501. https://doi.org/10.1088/1361-6528/ab4506.
  5. [5]A. Sagasti, V. Palomares, J.M. Porro, I. Orúe, M.B. Sánchez-Ilárduya, A.C. Lopes, J. Gutiérrez, Magnetic, magnetoelastic and corrosion resistant properties of (Fe–Ni)-based metallic glasses for structural health monitoring applications, Materials (Basel) 13 (2019) 57. https://doi.org/10.3390/ma13010057.
  6. [6]A. Sagasti, A.C. Lopes, A. Lasheras, V. Palomares, J. Carrizo, J. Gutierrez, J.M. Barandiaran, Corrosion resistant metallic glasses for biosensing applications, AIP Adv. 8 (2018). https://doi.org/10.1063/1.4994108.
  7. [7]L. Ren, K. Yu, Y. Tan, Wireless and passive magnetoelastic-based sensor for force monitoring of artificial bone, IEEE Sens. J. 19 (2019) 2096–2104. https://doi.org/10.1109/jsen.2018.2874943.
  8. [8]S. Shekhar, S.S. Karipott, R.E. Guldberg, K.G. Ong, Magnetoelastic sensors for real-time tracking of cell growth, Biotechnol. Bioeng. 118 (2021) 2380–2385. https://doi.org/10.1002/bit.27680.
  9. [9]R. Zhang, M. Tejedor, M. Anderson, M. Paulose, C. Grimes, Ethylene detection using nanoporous PtTiO2 coatings applied to magnetoelastic thick films, Sensors (Basel) 2 (2002) 331–338. https://doi.org/10.3390/s20800331.
  10. [10]M.K. Jain, C.A. Grimes, A wireless magnetoelastic micro-sensor array for simultaneous measurement of temperature and pressure, IEEE Trans. Magn. 37 (2001) 2022–2024. https://doi.org/10.1109/20.951041.
  11. [11]C.A. Grimes, S.C. Roy, S. Rani, Q. Cai, Theory, instrumentation and applications of magnetoelastic resonance sensors: A review, Sensors (Basel) 11 (2011) 2809–2844. https://doi.org/10.3390/s110302809.
  12. [12]X. Gao, W. Yang, P. Pang, S. Liao, Q. Cai, K. Zeng, C.A. Grimes, A wireless magnetoelastic biosensor for rapid detection of glucose concentrations in urine samples, Sens. Actuators B Chem. 128 (2007) 161–167. https://doi.org/10.1016/j.snb.2007.05.045.
  13. [13]K. Zeng, M. Paulose, K.G. Ong, C.A. Grimes, Frequency-domain characterization of magnetoelastic sensors: a microcontroller-based instrument for spectrum analysis using a threshold-crossing counting technique, Sens. Actuators A Phys. 121 (2005) 66–71. https://doi.org/10.1016/j.sna.2004.11.039.
  14. [14]K. Zeng, K.G. Ong, C. Mungle, C.A. Grimes, Time domain characterization of oscillating sensors: Application of frequency counting to resonance frequency determination, Rev. Sci. Instrum. 73 (2002) 4375–4380. https://doi.org/10.1063/1.1518128.
  15. [15]S. Sang, Y. Li, X. Guo, B. Zhang, X. Xue, K. Zhuo, C. Zhao, W. Zhang, Z. Yuan, A portable device for rapid detection of human serum albumin using an immunoglobulin-coating-based magnetoelastic biosensor, Biosens. Bioelectron. 141 (2019) 111399. https://doi.org/10.1016/j.bios.2019.111399.
  16. [16]W. Shen, Z. Zhang, S. Horikawa, A. Zhang, J. Teng, L.C. Mathison, B.A. Chin, Time domain characterization of magnetoelastic sensors: A pulse method for resonance frequency determination, Rev. Sci. Instrum. 81 (2010). https://doi.org/10.1063/1.3475153.
  17. [17]X. Yang, K. Ong, W. Dreschel, K. Zeng, C. Mungle, C. Grimes, Design of a wireless sensor network for long-term, in-situ monitoring of an aqueous environment, Sensors (Basel) 2 (2002) 455–472. https://doi.org/10.3390/s21100455.
  18. [18]Data Sheet, Low power, 12.65 mW, 2.3 V to 5.5 V, programmable waveform generator, Analog.com (2012). https://www.analog.com/media/en/technical-documentation/data-sheets/ad9833.pdf (acesso em 2 de maio de 2024).
  19. [19]Datasheet, LF–2.7 GHz RF/IF Gain and Phase Detector, Analog.com (2002). https://www.analog.com/media/cn/technical-documentation/data-sheets/AD8302.pdf (acesso em 2 de maio de 2024).
  20. [20]J. Labun, G. Kalapos, M. Ceskovic, P. Kurdel, J. Gamec, A. Nekrasov, C. Fidge, M. Lassak, A simple high-precision 2-port vector analyzer, IEEE Access 8 (2020) 196609–196617. https://doi.org/10.1109/access.2020.3034536.
  21. [21]B. Ben Atitallah, A.Y. Kallel, D. Bouchaala, N. Derbel, O. Kanoun, Comparative study of measurement methods for embedded bioimpedance spectroscopy systems, Sensors (Basel) 22 (2022) 5801. https://doi.org/10.3390/s22155801.
  22. [22]D.R. Santoso, B. Pitaloka, C.S. Widodo, U.P. Juswono, Low-cost, compact, and rapid bio-impedance spectrometer with real-time Bode and Nyquist plots, Appl. Sci. (Basel) 10 (2020) 878. https://doi.org/10.3390/app10030878.
  23. [23]B. Bergmair, T. Huber, F. Bruckner, C. Vogler, M. Fuger, D. Suess, Fully coupled, dynamic model of a magnetostrictive amorphous ribbon and its validation, J. Appl. Phys. 115 (2014). https://doi.org/10.1063/1.4861735.

Paper details

vlnc:W_zgv5bgpd
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 8

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents