Design and Modelling of a Modified High Gain Printed Vivaldi Antenna for EMC Measurements

The microstrip feed line width \(\:{w}_{f}\) for any impedance \(\:{z}_{0}\) can approximaltely be calculated from the following equation in [7].

For characteristic impedance \(\:{z}_{0}=50\:{\Omega\:}\), h = 1.6 mm and \(\:{\epsilon\:}_{reff}=4.2\) for FR-4 substrate, the microstrip line width is \(\:{w}_{f}\:\)= 3.03 mm, and this part is balanced since it will connect to the 50 Ω coaxial cable.

where, the effective relative permittivity is calculated based on the reported equation in [7].

Using the same equation, the characteristic impedance of the microstrip line connected to the unbalance part (antenna) is 90 \(\:{\Omega\:}\) at 1 mm width. Moreover, to match the two impedances (90 Ω and 50 Ω), a tapered line has been introduced to have a smooth transition and offer wide bandwidth [2]. The electromagnetic coupling feeding has been employed, which is based on the microstrip quarter-wavelength line/slot line transition power [22]. Suppose we have an input wave of \(\:\text{sin}wt\) as shown in Fig. 2a, the electromagnetic wave travels along the strip line and reverse its direction at the edge. The reflected signal will have a phase shift of 2π or zero at the center point as presented in [21] and as a result the phase shift will equal to 2π or zero at the center point presented in Fig. 2b.

The wideband behavior of impedance matching has been achieved by adopting stubs in microstrip and slot lines. These stubs have different shapes like radial stub and cavity slot, as shown in Fig. 3. The radial stub is an open-circuit stub that is used as an alternative for the straight quarter wavelength microstrip line for two reasons. Firstly, the radial stub offers a wideband short circuit wider than the linear quarter wavelength microstrip line. Secondly, at the low frequencies, the quarter wavelength microstrip line is significantly long and might not be very practical to implement with a straight quarter wavelength microstrip line.

To provide deep insight on the effect of radial stub in matching the proposed Vivaldi antenna with 50 Ω input port, two influencing factors (the radius R and angle ‘theta’ θ) are swept, optimized and analysis to study the effects of their variations on the impedance matching of the antenna. Figure 4 (a) presents the reflection coefficient versus frequency for different values of radial stub radius R. decreasing the length of the radial stub will increase the operational bandwidth up to 3 GHz at R = 9 mm. The next step is to change the radial stub angle θ with fixed value of R = 9 mm as shown in Fig. 4 (b). Increasing the value of θ will enhance the reflection coefficient at the lower frequency band at 1 GHz on the price of reduce the whole operational frequency range. The selected value of θ = 25⁰ is the optimum since it enhances the reflection coefficient at the lower frequency band and maintains on the same operational frequency range. The optimized values of electromagnetic coupling technique parameters are listed in Table 2.

This antenna has been designed and modeled in the CST Microwave Studio package based on the Finite Integration Technique (FIT) [23]. Figure 5(a) presents the simulated reflection coefficient versus frequency, while the simulated realized gain is shown in Fig. 5(b).

It can be seen that the reflection coefficient of the conventional design offers an acceptable bandwidth (0.85–5 GHz). On the other hand, the non-smoothy gain with low values through the whole band (in particular at the lower bands) would reflect an antenna factor with a very small margin for the compliance test. In fact, the antenna factor is one of the critical parameters to decide whether this antenna is good to use as a reference antenna in EMC measurement applications or not.

Two rectangular slots were engraved on both sides of the antenna close to the feeding point, which reflects better performance in reshaping the gain and making the transition smooth, as shown in Fig. 6. The optimum result of the realized gain is at slots length Ls1 = 55 mm, in which the antenna gain has a flatter response than other gain curves. For instance, when Ls1 = 70 mm, the deployed slot would be too close to the feeding point, and the realized gain decreases at the lower band at 2 GHz due to the electromagnetic coupling between the engraved rectangular slots and the feeding point.

Figure 7 demonstrates the optimum simulated realized gain and corresponding antenna factor versus frequency at slots length Ls1 = 55 mm. By observing the antenna factor results in Fig. 7(b), it has been found that the realized gain needs more enhancement to achieve an acceptable level of the antenna factor compared to the classical antenna. Therefore, another method is planned to be utilized to maximize the field and gain at the higher frequency bands in the next section.

The key to enhancing the antenna factor is manipulating the realized gain since they have a direct connection as in Eq. (4). Recently, several techniques have been proposed to enhance the antenna gain to suit many applications, like metamaterials [27, 28], reflectors [29, 30], superstrate [31, 32], and slots [26]. In particular, the realized gain of the Vivaldi antenna has also been enhanced with different techniques, like inserting a load resistor at the end of the coupling gap [33], using ultrathin microwave absorbing material at the edges of the antenna [34], and utilizing an index near zero meta-surface in the aperture [35]. In this work, two reversed triangular slots were added to the antenna’s top part, as shown in Fig. 8.

The position and the dimensions of the slots have been adjusted with the help of surface current distribution. Figure 9 presents the tangent component of the electrical field distribution with and without using two triangular slots. It is clear that the radiation at the side of the antenna is significantly reduced when engraving the two slots, so the radiation gets more directive in this case.

The triangular slots directly impact, reducing the radiation losses on both sides. Hence, the maximum radiation is in such a way that it is validated by observing the radiation pattern and maximum gain at 4.3 GHz, as demonstrated in Fig. 10.

As shown in Fig. 11 (a) and (b), the realized gain and the corresponding antenna factor, respectively, in the frequency range (2.4–6 GHz) are increased by more than 2 dB.

The positions of these slots have been determined using surface current distribution by looking for the electromagnetic waves that do not directly affect the resonance—cutting these parts results in an improvement in the realized gain because it makes the antenna radiation pattern to be more directive.

The fabrication process has been done in the EMC laboratory at the University of West Bohemia, Faculty of Electrical Engineering, with the help of a ProtoMat S-100 CNC machine. Figure 12(a) presents the fabricated structure of the final design. Furthermore, the S-parameters measurement is performed using RIGOL DSA875 Spectrum Analyzer stands up to 7.5 GHz, as shown in Fig. 12 (b). The simulation and measured results for the reflection coefficient as a function of frequency are shown in Fig. 13, which demonstrates an antenna operational band extends along the range (0.8–5.5 GHz).

The simulated and measured results for the realized gain and corresponding antenna factor are presented in Fig. 14 (a) and (b), respectively. The measured gain was achieved with the help of third antenna (reference antenna with known properties) inside the EMC chamber using the comparison method. This antenna offers good values of antenna factor compared to the commercial antenna, which makes it the right choice to use as a reference antenna.

Table 3 lists the achievements of the proposed technique in improving the realized gain compared to the previous articles. It can be seen that the technique used in the proposed structure tackles the aims of size reduction, bandwidth enhancements, and gain improvements. For instance, the proposed work has smaller size compared with ref [34], and ref [34] since they cover the lower frequency band to make fair comparison. Consequently, the proposed method has a considerable gain enhancement of about 3.3 dB at 4.3 GHz with maximum gain of 10.7 dBi compared to the reported literatures. It is worth mentioning that the use of two upside down rectangle slots techniques for gain improvements showed its simplicity and direct influence of enhancing the gain compared to the reported techniques

The balance or steady phase center is an obligatory specification for every reference antenna. It means that the radiation pattern transition with frequencies must be smooth, and the main-lobe position always stays in the same direction for the whole frequency band. The 3-D radiation patterns for different frequencies are presented in Fig. 15. It is evident that the antenna has a stable radiation pattern by noticing the maximum field’s transition with frequencies. Thus, the antenna can be qualified to serve as a reference antenna. On the other hand, co-polarization and cross-polarization are essential aspects related to the radiation pattern of the proposed antenna in both E-plane and H-plane since the structure is planned to be used as a reference antenna for EMC measurements. The radiation pattern measurement of this antenna has been performed inside the EMC chamber at the University of West Bohemia, Faculty of Electrical Engineering. The proposed antenna is mounted on the turntable and rotated around its axle in both vertical and horizontal direction to record the electric and magnetic field, respectively. “Moreover, the achieved maximum gain has horizontal direction that coincides with the directive Electrical field. Figure 16 shows the co-pol and cross-pol radiation patterns for E & H-plane at different frequencies like 1 GHz, 3 GHz, and 5 GHz. It can be seen that the cross-polarization level starts with small values and then increases with frequency due to the higher mode effect generated by this symmetrical structure. For instance, the cross-polarization component of the E-field begins with − 35 dB at 1 GHz and ends with − 25 dB at 5 GHz. At the same time, the cross-polarization of the H-field starts with − 30 dB at 1 GHz and grows to -15 dB at 5 GHz. It is worth noting that the increase in the cross polarization of the h-plane radiation is due to the undesired radiation modes with increasing frequency. However, the power level of the cross-polarization of the H-field (starts with − 30 dB at 1 GHz and − 15 dB at 5 GHz) does not affect the antenna performances.

The type of polarization is usually demonstrated by presenting an axial ratio (AR) value in dB. For instance, circular polarization has an axial ratio between 0 dB and 3 dB, an axial ratio with a value higher than 3 dB refers to elliptical polarization, and linear polarization is achieved when the AR value goes to infinity. The design of the EMC reference antenna is required to be linearly polarized. However, there will be some deviation between the designer’s goals and the actual realization. The transition region between the linear and elliptical polarization has some ambiguity since infinity is relative, and there is no practical or industrial form to distinguish the boundary between these two types. Furthermore, some references refer that linear polarization is a particular case of elliptical polarization. The proposed design in [38] has demonstrated linear polarization with AR > 10 dB, while authors in [39] have considered AR > 20 dB to be the boundary between linear and elliptical polarization. In this work, the axial ratio has a value of more than 20 dB for the whole frequency band, which indicates the linear polarization that perfectly satisfies the EMC requirement.

Most One of the critical parameters in EMC measurement is the vertical distance w that tangents the DUT, and this distance can be calculated using (6).

where d is the minimum measurement distance between the reference antenna and the DUT, which can be either 1 m, 3 m, or 10 m. Table 4 demonstrates the \(\:{\varnothing\:}_{3dB}\) beamwidth and its tangent distance w at minimum measurement distance d = 1 m for both Vivaldi antenna and the DRG Horn antenna of CISPR standard [40]. The half-power beamwidth angles for the proposed structure are agreeable with the CISPR standard.

A comparison between the specifications of the proposed antenna and the commercial antennas offered for sale [41] and [42] is listed in Table 5. The calibrated antenna PowerLOG^® PRO 30,800 has relative bandwidth of 180% with a size of 510 × 507 × 507 mm³ while EMC antenna type TBMA4 shows a relatively small size of 241 × 214 × 153 mm³ with relative bandwidth of 155%. In comparison, the proposed antenna offers a relative bandwidth of 150% with a compact and low-profile size of 180 × 165 × 1.6 mm³. The size reduction of the proposed antenna in both length and width are 64% and 67% compared to the PowerLOG^® PRO 30,800 antenna and 25% and 22% compared to the TBMA4 antenna. Furthermore, the gain of both structures is relatively good for EMC measurements reflecting good antenna factor values.

The numerical values of antenna factor versus frequency for the proposed structure, Ref [16], and the commercial antennas offered for sale are listed in Table 6. It is clear that the proposed antenna has a promising antenna factor making it a good candidate to serve as a reference antenna inside the EMC chamber.

The proposed structures in the referenced papers [2‐4] and [39] were compared with their equivalent classical structures to show their performance and their advantages in terms of size reduction, bandwidth enhancement, gain, antenna factor, and radiation pattern. Consequently, the achieved results in the proposed work are significant w.r.t references [2‐4] and [39]. For instance, the proposed antenna has a high gain of about 10.7 dBi compared to the cited papers, where the high value of gain reflects good antenna factor (in terms of stability and un uncertainty) and makes it applicable to use for radiated emission and radiated immunity test. Furthermore, the size reduction of this antenna is 64% and 67% compared to the classical horn antenna used for the same purpose. And finally, the use of two upside down rectangle slots techniques for gain improvements showed its simplicity and direct influence on enhancing the gain compared to the reported techniques in Table 3. It is worth mentioning that the proposed structure could be modified to exhibit a dual-polarized Vivaldi antenna for future work with appropriate calibration”.