The experiments described in this work were performed with the aim of introducing a new plasma antenna that was excited by a 5–20 kHz alternating current (AC) power supply, where the antenna was transformed into a U-shape. The results show that the impedance, voltage standing-wave ratio (VSWR), radiation pattern and gain characteristics of the antenna can be controlled rapidly by varying not only the discharge power, but also by varying the discharge frequency in the range from 5 to 20 kHz. When the discharge frequency is adjusted from 10 to 12 kHz, the gain is higher within a relatively broad frequency band and the switch-on time is less than 1 ms when the discharge power is less than 5 W, meaning that the plasma antenna can be turned on and off rapidly.
I. INTRODUCTION
Plasma antennas have attracted considerable research attention in recent years because they offer numerous distinct advantages over conventional metal antennas.1–4 The parameters of these antennas can be reconfigured conveniently by changing the discharge power, which makes it possible to use the antennas over a wide waveband and also enables simplification of the matching network. The plasma can be created and destroyed rapidly by simply applying an electrical pulse to the discharge tube, and thus plasma antennas can be switched on and off rapidly. When the antenna is on, it exhibits high conductivity, thus providing a conducting medium for the applied radio-frequency (RF) signal. When the antenna is off, it is non-conducting and does not interfere with any other devices. Plasma antennas can therefore play an integral and highly important role in many military and civil applications, including stealth and smart antennas. Given the growing importance and number of applications of plasma antennas, the characteristics of plasma antennas have already been studied for decades and many simulations of plasma antennas based on numerical methods and computational codes have been performed. The characteristics of interest include impedance, gain, phase, polarity, radiation, and dissipation, among others. Borg1,2 reported measurements of the efficiencies and the radiation patterns of plasma column antenna elements used in communications applications. Anderson and Alexeff4 successfully demonstrated the operation of plasma antennas in both transmission and reception modes. Moisans5 provided a review of surface-wave sustained plasma. Kumar and Bora6–8 have investigated the antenna properties of the different plasma structures of a plasma column for use as a reconfigurable plasma antenna. By varying the operating parameters, a single plasma antenna can be transformed into multiple antenna elements, which can then be arranged in different series. Russo and Cerri9–11 have researched the properties of the surfaguide-fed plasma antenna. The problems of the nonlinear behavior of plasma antenna vibrators have been studied by Belyaev et al.12 Zhang et al.13 investigated the interactions between electromagnetic waves and glow plasma. Lee,14 Qian15 and Li16,17 have all presented numerical analyses of plasma-column antennas using the finite-difference time-domain (FDTD) method.
However, there are some problems that are yet to be solved. Over the past few decades, 50 or 60 Hz AC power supplies were used to produce the plasmas, however, these power supplies were found to produce very high levels of noise.3 For the vast majority of plasma antennas, the plasma is produced using high frequency (≥1 MHz, HF) power supplies. These antennas are usually called RF plasma antennas or surface wave plasma antennas, and have been shown to maintain a specific gain and low noise.18 However, the bandwidths of such antennas for the pumping frequency are limited, and the cost of the pumping RF generator is also high. In addition, the problem of the simultaneous presence of two RF signals in the excitation and signal channels must also be considered.11 The problems of electromagnetic interference and bandwidth requirements must therefore be solved urgently. In common with many other plasma sources, a 5–20 kHz AC power supply is used in this work, and presents some advantages in the design of plasma antennas: no bandwidth limitations, low excitation power, low noise and simple realization. In addition, use of this power supply also allows the complex task of reducing the strong coupling between the pump and the radiated signals to be avoided. Most importantly, we can not only change the discharge power, but can also vary the discharge frequencies to alter the characteristics of the plasma antenna.
II. ANTENNA STRUCTURE AND TESTING
A commercially available tube with an inner diameter of 10 mm, an outer diameter of 12 mm and a length of 1200 mm that could be filled with argon, neon, argon with mercury, or neon with mercury at different pressure levels was used in this work. The structure of the plasma antenna is shown in Fig. 1(a); the tube is U-shaped, with two electrodes inserted in an insulating box in which the 5–20 kHz AC adjustable power supply was installed. A capacitive coupling scheme was used for the signal coupling system. The two capacitive couplers were made from copper, with a width of 30 mm; they were placed at each side of the tube and were shielded by a well-sealed shielding box made from cast aluminum that was connected to ground to reduce any electromagnetic (EM) radiation from the copper couplers. One coupler was connected to the transmission line to apply useful signals and the other was connected directly to ground. The distance between the centers of these couplers and the electrodes of the discharge tube was at least 200 mm to prevent the high AC voltage from damaging the instrument. The plasma column was created by applying the 5–20 kHz AC source and the gas was ionized by a strong electric field that was applied at the ends of the plasma antenna. It was easier for the U-shaped plasma antenna to be connected to the power supply than a linear antenna. In addition, the U-shape allowed the antenna to be smaller in size. We also adopted another structure, as shown in Fig. 1(b); unlike Fig. 1(a), only one capacitive coupler was connected to the transmission line, and thus the cast aluminum box connected to ground was only half the size of that shown in Fig. 1(a).
The structure and the working programs of the 5–20 kHz power supply are shown in Fig. 2. The input voltage was 220 V AC at a frequency of 50 Hz, and this voltage was sent to the AC/DC bridge-type full wave rectifier circuit after passing through a filter. Consequently, the 50 Hz AC voltage was converted into a pulsating DC voltage, and then into a smoothed form with a voltage of about 300 V; finally, the DC voltage was converted into the required 5–20 kHz AC output by the half-bridge converter. The plasma that was generated by the 5–20 kHz power supply was diagnosed using a Langmuir probe at different discharge powers and at discharge frequencies in the range from 5 to 20 kHz. The probe is 0.1 mm in diameter and is made from tungsten. To reduce measurement errors, the pressure of the gas used to fill the discharge tube is no more than 500 Pa.
In Fig. 1(a), it was clear that the structure was similar to that of a loop antenna, and it was a complex structure when compared with that of Fig. 1(b). The plasma antenna was thus established according to the structure shown in Fig. 1(b), which was similar to that of the linear plasma antenna, as shown in Fig. 3. The full length of the discharge tube, which was made from toughened glass, was 1200 mm, and it was filled with argon and mercury. Only one capacitive coupler was installed in the shielding box, and the coupler was connected to the transmission line through the RF port. A vacuum device was used to adjust the gas pressure. To test the characteristics of this 5–20 kHz plasma antenna, the antenna was placed in an anechoic chamber. The voltage standing-wave ratio (VSWR), gain and direction characteristics were studied using a network analyzer with output impedance of 50 Ω, which was equal to the characteristic impedance of the transmitting line. We also established a 40.68 MHz AC linear plasma antenna and a metal linear antenna. In the 40.68 MHz AC plasma antenna system, unlike the 5–20 kHz AC plasma antenna, the excitation end was in the form of a capacitive coupler outside the discharge tube, and not a hollow electrode inserted into the tube. A model of the 40.68 MHz AC plasma antenna is shown in Fig. 4.
Plasma behaves like a dielectric with a permittivity of less than unity at frequencies above the plasma frequency.14 In this case, EM waves can propagate through the plasma in accordance with the dispersion relationship , where ωp is the plasma frequency and k is the wave number. At frequencies below the plasma frequency, where the real part of the permittivity of the plasma is negative, the plasma behaves like a waveguide below cutoff, in that electromagnetic waves do not propagate in the radial direction. In contrast, in the axial direction, a surface wave propagates along the plasma column. When the plasma frequency increases, the attenuation coefficient decreases or approaches zero. Therefore, when the plasma frequency is far higher than the signal frequencies, the propagation characteristics of the EM waves in a plasma antenna are similar to those in a metal antenna.
III. RESULTS AND ANALYSIS
A. Discharge properties of power supply
An oscilloscope was used to measure the discharge voltage and current. A high voltage probe (HP-60, Iwatsu, Iwasaki, Japan) was connected between the oscilloscope and the high voltage electrode, and a resistor was connected between another electrode and ground to test the current according to Ohm’s law for AC circuits. To make the circuit have a pure resistance characteristic, a tuning circuit was required. Figure 5 shows the voltage and current characteristics of the plasma antenna working at discharge frequencies of (a) 10 kHz and (b) 20 kHz. The excitation voltage was unchanged at 560 V and the filled gas was an argon and mercury mixture at a gas pressure of 500 Pa, while the phase difference between the voltage and the current was nearly zero. The discharge current decreased from 22 to 8 mA with an increase in the frequency of the power supply from 10 to 20 kHz. The average discharge powers of the 10 and 20 kHz discharge frequencies were both less than 10 W, as measured using the power meter. The power was also calculated in the form of half the peak voltage multiplied by the peak current, and the results were almost the same as those measured by the power meter. Table I shows the maximum currents and powers at the different discharge frequencies of the power supply.
B. Electron density
Several Langmuir double probes made from tungsten, all with a diameter of 0.1 mm, were inserted along the axial direction of the plasma antenna. We were thus able to measure the electron density at several positions. The plasma electron densities for the different discharge power levels, ranging from 2 to 12 W, are shown in Fig. 6(a). In these experiments, the discharge tube was filled with argon at a gas pressure of 500 Pa. The results show that the density increases with increasing discharge power, varying from 1016 m−3 at approximately 1 W to nearly 1017 m−3 above 6 W. We measured the plasma parameters at several different locations along the plasma antenna using the Langmuir probes. The electron temperature and density did not change much along the axial direction. Figure 6(b) shows the electron density for various frequencies from 5 to 20 kHz. The plasma density gradually decreased from nearly 1017 m−3 in the 6 to 12 kHz range to approximately 1016 m−3 at 20 kHz, using an unchanged excitation voltage of 560 V, gas pressure of 500 Pa and a gas composition of argon mixed with mercury. We used the 40.68 MHz AC power supply to generate the plasma antenna, because its discharge power was much higher than that of the 5–20 kHz power supply. While the 40.68 MHz AC power supply can generate a density of 1017 m−3, the discharge power must be very high (reaching about 50 W in the experiments) to generate and maintain the plasma.
Electron density of different discharge conditions (a) different discharge power P0, (b) different discharge frequency f0.
Electron density of different discharge conditions (a) different discharge power P0, (b) different discharge frequency f0.
C. Switch-on time
Another goal of our investigation was to pursue an approach involving the switch-on time of the plasma antenna, which reflected the established velocity of the plasma. In the experiments, a photomultiplier tube and an oscilloscope were used to estimate the switch-on time of the plasma using the light-electricity time difference method that was invented as part of our previous work.19 Figure 7 shows a schematic diagram of the setup used to measure the switch-on time of the 5–20 kHz AC plasma antennas. According to the procedure of Ref. 19, two channels were connected to the oscilloscope, where one was used for the discharge signals and the other was used for the optical signals. To protect the oscilloscope from the high voltage conditions, a voltage divider (2000:1) was inserted between the electrode and the port of the oscilloscope. The principle of the measurement was simple, i.e., we used the trigger principle of the oscilloscope.
Schematic of experimental setup for measurement method of light-electricity time difference for measuring the switch-on time of 5-20 kHz ac plasma antenna. 1 plasma antenna, 2 gas tank, 3 vacuum pump, 4 oscilloscope, 5 photomultiplier, 6 voltage divider.
Schematic of experimental setup for measurement method of light-electricity time difference for measuring the switch-on time of 5-20 kHz ac plasma antenna. 1 plasma antenna, 2 gas tank, 3 vacuum pump, 4 oscilloscope, 5 photomultiplier, 6 voltage divider.
To provide a better understanding of this experimental study, the switch-on time ton can be written as:
where ta is the optical signal’s initial time from the plasma, tb is the electrical signal’s initial time from the power supply, tl is the response time of the photomultiplier tube and tr is the propagation time of the plasma optical signals from the plasma to the spectrum probe. In the experiments, tr and tl were very small and could thus be ignored, so the switch-on time could therefore be expressed approximately as
The measured switch-on time of the plasma antenna operating at 12 kHz is shown in Fig. 8. In Fig. 8, the upper curve is the optical signal received by the photomultiplier and the lower curve is the discharge voltage. We observed experimentally that when the plasma antenna was filled with argon alone, the switch-on time was approximately 1 ms.
D. Radiation characteristics
In the 5–20 kHz plasma antenna system, unlike the 40.68 MHz AC system, there was no strong coupling between the excitation and signal channels. We proved this opinion using the network analyzer (Agilent E5071C, California, US, 9 kHz-4.5 GHz), as shown in Fig. 9, and the excitation and signal channels of the two power supplies were very clearly visible. Then, the echo loss S11 and the coupling factor S21 at the frequencies of 10 kHz and 40.68 MHz were studied experimentally. The S11 value at 10 kHz was nearly zero, and the value at 40.68 MHz was approximately −4 dB, as shown in Table II. In addition, S21 was shown to be about −60.4 dB at 10 kHzand −32.5 dB at 40.68 MHz. Therefore, no strong coupling existed between the excitation and signal channels. We could thus measure the input impedance and the VSWR directly using the network analyzer without any protection circuitry. The output impedance of the analyzer was 50 Ω, which was equal to the characteristic impedance of the transmission line.
Schematic diagram of research on coupling between excitation and signal channels.
Schematic diagram of research on coupling between excitation and signal channels.
The S11 and S21 of plasma antenna at frequencies of 10 kHz and 40.68 MHz.
Frequency . | S11 (dB) . | S21 (dB) . |
---|---|---|
10 kHz | 0 | −4 |
40.68 MHz | −60.4 | −32.5 |
Frequency . | S11 (dB) . | S21 (dB) . |
---|---|---|
10 kHz | 0 | −4 |
40.68 MHz | −60.4 | −32.5 |
In the measurements, the discharge power was fixed at 6 W and the frequency was varied from 5 to 20 kHz. Figure 10 shows the VSWR results for the 5–20 kHz system when filled with argon and mercury at a pressure of 300 Pa. As shown, the VSWR of the plasma antenna when working at the discharge frequency of 10 kHz was less than 2 at frequencies between 183 and 340 MHz. We estimated the relative frequency bands of the antennas using the formula
where ϵ is the impedance bandwidth factor of the researched device (i.e., the plasma antenna), and fhigh and flow are the highest and lowest frequencies of the researched range, respectively, when the VSWR was less than 2. Accordingly, the impedance wideband factor ϵ was equal to 0.3. We adjusted the discharge frequency to a value of 12 kHz, and as shown in Fig. 10, the VSWR was less than 2 in the signal frequency range from 183 to 500 MHz. The impedance bandwidth factor was equal to 0.46. If the antenna frequency band was in the range from 0.25 to 1, the characteristics were such that the systems could be termed “wideband” or “ultra-wideband” systems. Simultaneously, the pass-band of the plasma antenna was estimated via a conventional method using the equation
where Δf = fhigh − flow, and f0 is the center frequency of the researched range. Therefore, the pass bands were 60% at 8 kHz and 92% at 12 kHz.
The discharge frequency was then adjusted to 15 kHz, and the waveband (i.e., the frequency band where the VSWR was less than or equal to 2) moved to the right, from 280 to 650 MHz. The plasma antenna would not work well in the frequency band ranging from 100 to 400 MHz with further increases in the discharge frequency. Therefore, when the discharge frequency increases, the waveband moves to the right. It was also clear that the plasma antenna had a wideband impedance characteristic and that the frequency band could be changed by adjusting the discharge frequencies.
E. Antenna gain
The gain of the plasma antenna must also be considered. The network analyzer was also used in these experiments to test the gain (S21) values by varying the discharge frequencies and the filled gas (neon, argon, neon with mercury, and argon with mercury). Previously, we had shown that in a certain gas pressure range, the antenna gain increased with increasing pressure. Figure 11 depicts the gains of the plasma antenna when it is filled with the different inert gases, i.e., neon, argon, neon with mercury, and argon with mercury, at a gas pressure of 500 Pa and with a discharge frequency of 12 kHz. The test for S21 was similar to that shown on the left of Fig. 9. The VSWR of the experimental bandwidth was less than 2, and the trends of the gain of the plasma antenna when filled with the four different types of gases were nearly the same. In addition, the gain was very low for the antenna filled with neon because of its relatively low degree of ionization, while plasma antennas filled with argon or argon with mercury could maintain relatively ideal gains in a specific bandwidth. The results of these experiments were slightly different from those of Ref. 14. First, impedance matching was considered in our experiment; when the signal frequency was less than 150 MHz, the VSWR was higher than 2, meaning that the majority of the signal power was reflected back towards the signal generator, and the signal power received by the receiving antenna was thus lower. Second, in our experiments, the plasma density was close to 1017 m−3 and the plasma frequency fp was approximately 2700 MHz, which was beyond the research scope of Ref. 14.
Gains of plasma antenna versus frequency for different filling gases.
The S21 characteristics of the plasma antenna when filled with argon with mercury at about 500 Pa are shown in Fig. 12, where the discharge power was fixed at 6 W, and the discharge frequencies were 7, 12, 14 and 18 kHz. It was obvious that the S21 characteristics of plasma antennas working at 7 and 12 kHz were higher than the characteristics of those working at 14 kHz and 18 kHz. The plasma conductivity σ could be expressed as
where e is the electron charge, ne is the electron density, me is the electron mass, ω is the angular frequency of the electromagnetic wave, and ven is the electron-neutron collision frequency. At a fixed collision frequency, the conductivity increases with increasing electron density. Electron density was lower for discharge frequencies above 14 kHz. Therefore, the gain was relatively low. However, the gain bandwidth was relatively narrow for discharge frequencies of less than 9 kHz. The average gain was higher over a wider frequency band on the condition that the discharge frequency was in the range from 10 to 12 kHz.
Gains of plasma antenna versus frequency for different RF discharge frequencies.
Gains of plasma antenna versus frequency for different RF discharge frequencies.
F. Radiation pattern
Another aspect of our investigation involved the radiation pattern of the plasma antenna. Figure 13 shows the normalized radiation pattern of a plasma antenna filled with an argon and mercury mixture and working at a discharge frequency of 12 kHz with power of 6 W. The plasma antenna was placed in an anechoic chamber, so that the receiving antenna did not receive any spurious signals that were reflected from the surroundings in addition to those from the plasma antenna, which was a transmitting antenna. The incident signals were varied from 160 to 250 MHz for a VSWR of less than 2. Figure 13(a) shows the vertical radiation patterns at the signal frequencies of 160, 200 and 250 MHz, with VSWRs of 1.27, 1.14 and 1.07, respectively. The radiation patterns at the three frequencies were obviously intensely similar in the vertical and horizontal planes; the patterns of the plane were omnidirectional, similar to the whip antenna, and the shapes of the vertical planes were bidirectional. There was little difference among the receiving powers at the three signal frequencies, and the directions for the maximum powers were approximately 90° and 270°. Figure 13(b) shows the vertical and horizontal planes of the plasma antenna with a discharge frequency of 12 kHz, a signal frequency of 200 MHz, a gas mixture of argon with mercury, and an operating pressure of 500 Pa. The plasma antenna shows excellent bandwidth characteristics. We also compared the 5–20 kHz plasma antenna to the 40.68 MHz antenna, and while there was little difference between the gain and noise results for the two plasma antennas in the range from 30 to 200 MHz, the discharge power of the 40.68 MHz plasma antenna was higher than that of the 5–20 kHz antenna. More results from this study will be presented as part of our future work.
Radiation pattern of plasma antenna driven by 5-20 kHz ac power supply.
IV. CONCLUSION
A new type of U-shaped plasma antenna is established using a 5–20 kHz AC power supply. The antenna shows promising results in that it can not only be reconfigured by adjusting the discharge power, but it can also be changed by adjusting the discharge frequency.
The plasma antenna is not limited in terms of bandwidth and has been proved to have lower noise performance than 50 Hz AC and DC plasma antennas. The impedance and gain bandwidth can be changed rapidly by adjusting the discharge parameters, such as the frequency. We can also obtain high gains to make the proposed plasma antenna feasible for communication applications when using gas components of argon or argon with mercury, and operating at a gas pressure in the range from 300 to 500 Pa, a discharge frequency in the range from 10 to 12 kHz and a discharge power of approximately 6 W. The switch-on time is less than 1 ms, meaning that the plasma antenna can be turned on and off rapidly. In addition, the 5–20 kHz plasma antenna shows excellent bandwidth characteristics, and the cost of the 5–20 kHz power supply is much lower than that of the HF AC supply.
ACKNOWLEDGMENTS
The authors gratefully acknowledge the support of the National Natural Science Foundation of China (Grant no. 51279099), the Shanghai Municipal Science and Technology Commission Project (Grant no. 13510501600), the Development Fund for Shanghai Talents (Grant no. 201436) and China Postdoctoral Science Foundation (Grant no. 2015M581585).