Probe Fed Stacked Patch Antenna for UWB Sectoral ...
Probe Fed Stacked Patch Antenna for UWB Sectoral Applications Hassan Ghannoum, Serge Bories and Christophe Roblin, Member, IEEE ENSTA UEI 32 Bd. Victor 75739 PARIS FRANCE ([email protected])
Abstract— A microstrip patch antenna with two E-shaped stacked patches for UWB sectoral applications is proposed in this paper. The E-shaped patch antenna has an impedance bandwidth of about 34%. By adding a second E-shaped patch at the top of the first patch a bandwidth of 54% has been obtained. The characteristic dimensions of the second patch as well as the shift between the two patches have been optimized to achieve the ultrawide bandwidth and a radiation pattern stability over the whole band. The distorting nature of this antenna has been quantified using time domain characterization tools and the influence of the ground plane on impedance bandwidth and radiation has been studied. Index Terms— UWB antennas, Stacked Microstrip Antennas, E-shaped patch, Time Domain Characterization Tools, Antenna Distortion.
M
I. INTRODUCTION
icrostrip patch antennas have several well-known advantages, such as low profile, low cost, light weight, ease of fabrication and conformity. However, the main disadvantage that limits the use of patch antennas in UWB communications is their narrow bandwidth, due to their inherent nature as resonant devices. To overcome this problem many efforts have been made and many methods have been proposed in the literature, including the use of thick substrates, cutting slots suitably in the metallic patch and introducing parasitic patches either on the same layer or on the top of the main patch. Recently, aperture-coupled fed stacked patch antenna [1] have been investigated and bandwidths up to 69% have been reported, however, the major drawbacks are the level of back radiation due to the use of a resonant aperture and the surface wave excitation. Other feeding techniques such as the use of L-shaped or F-shaped probes have also been proposed yielding to wide impedance bandwidths [2, 3], at the expense of increased complexity of the design and fabrication, especially of the probe. In [2], an L-probe fed stacked U-Slot patch antenna was proposed with a bandwidth up to 44.4% being achieved. V-slotted rectangular microstrip antenna with a stacked patch has been shown able to achieve bandwidths as high as 47% [4]. In this paper, we would like to propose an UWB stacked patch antenna designed by combining simultaneously several
methods among those proposed to enlarge the bandwidth of patch antennas, while maintaining a simple feed configuration. The proposed antenna is based on an E-shaped patch [5] coupled to another E-shaped patch stacked on the top of the first patch [6] achieving 54% input impedance bandwidth. Besides large impedance matching and beamwidth, for UWB sectoral applications, the behavior of the antenna in its main lobe has also to be investigated. Thus, the distorting nature of the proposed antenna has been quantified using characterization tools in the time domain (TD). The influence of the ground plane size on impedance bandwidth and gain has also been studied. II. ANTENNA STRUCTURE The proposed antenna structure is shown in Fig. 1. The antenna is made of two E-shaped stacked patches, two foam layers, and a vertical probe connected to the lower patch. The lower E-shaped patch is based on the design proposed in [5]. The design of this patch was deduced from a U-shaped patch by removing the portion with low current distribution. The lower E-shaped patch was designed for a [3.1-4.3] GHz bandwidth and the height of the first foam layer was around 7 mm. The upper E-shaped patch is based on the design proposed in [6]. Slots play an important role to control the wide-band behavior of the E-shaped patch. By only adjusting the length, width, and position of the slots, one can obtain satisfactory performances. In addition to the second foam layer thickness, slots positions, lengths and widths, an additional parameter had to be optimized: the offset distance (S) (See Fig. 1) between the patches. This latter parameter was found to affect the input impedance bandwidth as well as the radiation pattern especially at higher frequencies range. It affects the way the two patches couple to each other. The optimized value for the offset distance and the characteristic dimensions of the two E-shaped patches are shown in Table 1. A 12 mm foam layer was used as the second layer. It should be noted that the foam used has a permittivity of 1.08. Due to the thickness of the first foam layer (7.2 mm), the probe inductance was found to be relatively high when low diameter probes were used. Use of a washer on the probe aids to cancel the reactance of the probe at the expense of increased complexity of the design and fabrication. Thus, a 4 mm diameter probe was used with a conical transition to 0.9 mm diameter at the end for 50Ω SMA connector compatibility.
H plane
E plane
(a) Lower E-shaped patch
Fig. 2. Prototype of the antenna. TABLE I PHYSICAL DIMENSIONS OF THE STACKED ANTENNA (Dimensions unit: mm) P1 W2 Ls1 17.3 5.75 23.8 P2 6.35
(b) Upper E-shaped patch and (c) probe
Ps2 6.3
Ws2 1.325
W1 15.15
L1 20.9
Ws1 4
W3 2.875
Ls2 17.8
h1 7.2
h2 12
S 13
III. PERFORMANCES A. Technical Approach The MoM method tool WIPL-D [7] has been used for simulations. The return loss and the radiation pattern were measured in an anechoic chamber with an HP8510C® vector network analyzer and a calibrated 1-18 GHz Log Periodic Dipole Array (LPDA) antenna [8]. The following results are de-embedded so that they only show the antenna behaviour.
(d) Top view and (e) side view Fig. 1. The geometry of the stacked antenna and its dimensions (units: millimeters).
B. Frequency Domain Results The offset (S) between the two patches plays an important role in controlling the input impedance bandwidth as well as the radiation pattern at higher frequencies. Fig. 2 shows this effect on the antenna characteristics. When the upper patch is cantered at the top of the first one, the antenna has a bandwidth as large as the bandwidth of the lower patch alone. At this position the upper patch is not well coupled to the lower one. Through the study of the current distribution on the lower patch at higher frequencies, we concluded that the upper patch had to be shifted towards x > 0 in order to be properly excited.
As shown in Fig. 3 for S = 13 mm the upper patch is well coupled to the lower one and a 54% bandwidth is obtained. In fact, the upper patch is well coupled and the bandwidth remains practically unchanged for S between 10 mm and 14 mm, but the radiation pattern at higher frequencies depends on the choice of S in this interval. S = 13 mm was found to be the optimal value and the prototype was fabricated with respect to this offset value.
(b) Fig. 4. (a) Measured (blue) and simulated (red dash) return loss (dB), and (b) effective gain.
(a) Fig. 3. Calculated S11 (dB) with different shift values: 0 mm (blue, circle), 3 mm (yellow, square), 8 mm (magenta, point), 13 mm (green, plus).
The measured return loss for the proposed antenna is presented in Fig. 4. The measured input bandwidth with respect to S11
Abstract— A microstrip patch antenna with two E-shaped stacked patches for UWB sectoral applications is proposed in this paper. The E-shaped patch antenna has an impedance bandwidth of about 34%. By adding a second E-shaped patch at the top of the first patch a bandwidth of 54% has been obtained. The characteristic dimensions of the second patch as well as the shift between the two patches have been optimized to achieve the ultrawide bandwidth and a radiation pattern stability over the whole band. The distorting nature of this antenna has been quantified using time domain characterization tools and the influence of the ground plane on impedance bandwidth and radiation has been studied. Index Terms— UWB antennas, Stacked Microstrip Antennas, E-shaped patch, Time Domain Characterization Tools, Antenna Distortion.
M
I. INTRODUCTION
icrostrip patch antennas have several well-known advantages, such as low profile, low cost, light weight, ease of fabrication and conformity. However, the main disadvantage that limits the use of patch antennas in UWB communications is their narrow bandwidth, due to their inherent nature as resonant devices. To overcome this problem many efforts have been made and many methods have been proposed in the literature, including the use of thick substrates, cutting slots suitably in the metallic patch and introducing parasitic patches either on the same layer or on the top of the main patch. Recently, aperture-coupled fed stacked patch antenna [1] have been investigated and bandwidths up to 69% have been reported, however, the major drawbacks are the level of back radiation due to the use of a resonant aperture and the surface wave excitation. Other feeding techniques such as the use of L-shaped or F-shaped probes have also been proposed yielding to wide impedance bandwidths [2, 3], at the expense of increased complexity of the design and fabrication, especially of the probe. In [2], an L-probe fed stacked U-Slot patch antenna was proposed with a bandwidth up to 44.4% being achieved. V-slotted rectangular microstrip antenna with a stacked patch has been shown able to achieve bandwidths as high as 47% [4]. In this paper, we would like to propose an UWB stacked patch antenna designed by combining simultaneously several
methods among those proposed to enlarge the bandwidth of patch antennas, while maintaining a simple feed configuration. The proposed antenna is based on an E-shaped patch [5] coupled to another E-shaped patch stacked on the top of the first patch [6] achieving 54% input impedance bandwidth. Besides large impedance matching and beamwidth, for UWB sectoral applications, the behavior of the antenna in its main lobe has also to be investigated. Thus, the distorting nature of the proposed antenna has been quantified using characterization tools in the time domain (TD). The influence of the ground plane size on impedance bandwidth and gain has also been studied. II. ANTENNA STRUCTURE The proposed antenna structure is shown in Fig. 1. The antenna is made of two E-shaped stacked patches, two foam layers, and a vertical probe connected to the lower patch. The lower E-shaped patch is based on the design proposed in [5]. The design of this patch was deduced from a U-shaped patch by removing the portion with low current distribution. The lower E-shaped patch was designed for a [3.1-4.3] GHz bandwidth and the height of the first foam layer was around 7 mm. The upper E-shaped patch is based on the design proposed in [6]. Slots play an important role to control the wide-band behavior of the E-shaped patch. By only adjusting the length, width, and position of the slots, one can obtain satisfactory performances. In addition to the second foam layer thickness, slots positions, lengths and widths, an additional parameter had to be optimized: the offset distance (S) (See Fig. 1) between the patches. This latter parameter was found to affect the input impedance bandwidth as well as the radiation pattern especially at higher frequencies range. It affects the way the two patches couple to each other. The optimized value for the offset distance and the characteristic dimensions of the two E-shaped patches are shown in Table 1. A 12 mm foam layer was used as the second layer. It should be noted that the foam used has a permittivity of 1.08. Due to the thickness of the first foam layer (7.2 mm), the probe inductance was found to be relatively high when low diameter probes were used. Use of a washer on the probe aids to cancel the reactance of the probe at the expense of increased complexity of the design and fabrication. Thus, a 4 mm diameter probe was used with a conical transition to 0.9 mm diameter at the end for 50Ω SMA connector compatibility.
H plane
E plane
(a) Lower E-shaped patch
Fig. 2. Prototype of the antenna. TABLE I PHYSICAL DIMENSIONS OF THE STACKED ANTENNA (Dimensions unit: mm) P1 W2 Ls1 17.3 5.75 23.8 P2 6.35
(b) Upper E-shaped patch and (c) probe
Ps2 6.3
Ws2 1.325
W1 15.15
L1 20.9
Ws1 4
W3 2.875
Ls2 17.8
h1 7.2
h2 12
S 13
III. PERFORMANCES A. Technical Approach The MoM method tool WIPL-D [7] has been used for simulations. The return loss and the radiation pattern were measured in an anechoic chamber with an HP8510C® vector network analyzer and a calibrated 1-18 GHz Log Periodic Dipole Array (LPDA) antenna [8]. The following results are de-embedded so that they only show the antenna behaviour.
(d) Top view and (e) side view Fig. 1. The geometry of the stacked antenna and its dimensions (units: millimeters).
B. Frequency Domain Results The offset (S) between the two patches plays an important role in controlling the input impedance bandwidth as well as the radiation pattern at higher frequencies. Fig. 2 shows this effect on the antenna characteristics. When the upper patch is cantered at the top of the first one, the antenna has a bandwidth as large as the bandwidth of the lower patch alone. At this position the upper patch is not well coupled to the lower one. Through the study of the current distribution on the lower patch at higher frequencies, we concluded that the upper patch had to be shifted towards x > 0 in order to be properly excited.
As shown in Fig. 3 for S = 13 mm the upper patch is well coupled to the lower one and a 54% bandwidth is obtained. In fact, the upper patch is well coupled and the bandwidth remains practically unchanged for S between 10 mm and 14 mm, but the radiation pattern at higher frequencies depends on the choice of S in this interval. S = 13 mm was found to be the optimal value and the prototype was fabricated with respect to this offset value.
(b) Fig. 4. (a) Measured (blue) and simulated (red dash) return loss (dB), and (b) effective gain.
(a) Fig. 3. Calculated S11 (dB) with different shift values: 0 mm (blue, circle), 3 mm (yellow, square), 8 mm (magenta, point), 13 mm (green, plus).
The measured return loss for the proposed antenna is presented in Fig. 4. The measured input bandwidth with respect to S11
