Multi-purpose reconfigurable impedance matching network designs and antenna applications

Abstract

Modern radio frequency communication front-ends utilize multiple parallel and cascaded transceivers to accommodate new communication standards that require additional frequency band allocations. However, the utilization of more and more transceiver modules to accommodate emerging communication standards results in higher costs, additional power consumption, larger circuit board sizes, and an increased probability of failure. Reconfigurable radio frequency circuits have become a promising solution to eliminate the deployment of multiple transceivers. A reconfigurable transceiver can cover multi-ple communication standards by changing its operational frequency that conventional front-ends require numerous transceivers. Also, reconfigurable transceivers can be configured with new communication standards without necessitating new transceiver designs. The use of reconfigurable circuit designs is not only advantageous in terms of reducing costs, board sizes, and enabling versatility in the utilization of new standards but also performance deviations caused by external effects and damaged parts of the circuits can be compensated. Owing to the mentioned advantages, interest in reconfigurable circuits has increased gradually both in academia and industry. The replacement of conventional circuits with reconfigurable versions has been studied in literature and basic implementations have been utilized by the industry over the years. One of these circuits which is used in all front-ends is impedance matching networks. Since impedance matching networks are at the core of all radio frequency and microwave circuit designs, reconfigurable versions have to be deeply investigated. In this thesis, novel frequency and load-tunable multi-purpose reconfigurable impedance matching network designs are proposed by starting from their building blocks and designed networks are evaluated in terms of various performance metrics. Microstrip transmission lines that are used in the reconfigurable impedance matching networks are simulated and measured first. Electromagnetic (EM) simulations and measurements are in good agreement and indicate less than 0.5 dB insertion loss with less than 1% signal power reflection inside the 0.1 GHz-4.5 GHz frequency band. Since the reconfigurable networks include tuning elements that have to be properly controlled without introducing additional power loss, a DC bias network is designed and examined. The bias network includes a DC blocking capacitor, two decoupling capacitors, and an RF choke inductor. Component values are calculated by considering operational frequency. Schematic, EM Co-simulations, and measurements revealed that insertion loss of the bias network is less than 0.25 dB with -20 dB return loss at the input and -30 dB isolation at the DC port. Also, the designed bias network achieves complete DC blocking performance under 8 V which is the control voltage limit of the tuning components. In this way, tunable circuit elements that are sensitive to DC voltage change can be controlled independently from each other. As a final building block, hyperabrupt varactor diodes are examined with analytical calculations, simulations, and measurements. Since hyper-abrupt varactor diodes are more sensitive to DC bias voltage than conventional abrupt varactor diodes, their equivalent circuit models are less accurate. In order to construct a more accurate measurement-based model, first, the low-frequency response of the varactor diodes is examined. The grading factor and capacitive tuning ratio of the varactor diode are corrected according to measurement results. High frequency response of the varactor diode demonstrates self resonance frequency which is used for revision of package parasitics. Finally, the varactor diode’s Q-factor and insertion loss are obtained to evaluate applicability on the desired multi-purpose reconfigurable impedance matching network. The Q-factor of the varactor diode increases with the applied reverse bias voltages and reaches to 1100 at 50 MHz. Insertion loss is less than 0.35 dB between 0.1 GHz and 5 GHz. Both the Q-factor and insertion measurements reveal that the selected varactor diode suits expectations on the tuning element well. After examining building blocks, and constructing measurement-based circuit models, a reconfigurable impedance matching network is analytically synthesized with and without including varactor diode parasitics. The initial design is then modified to reach a better Q-factor which includes fewer components on the series branch of the circuit. A modified design is also synthesized with and without parasitic injection. First, designed matching networks are examined in terms of insertion loss with schematic, EM Co-simulations, and measurements. Schematic simulations provide better estimation for measurement results in initial designs whereas EM Co-simulations are in better agreement with measurements for modified designs. The initial and modified designs with parasitic injection have less than 1.5 dB insertion loss and less than 2.5 dB insertion loss without parasitic injection inside the operational frequency band. Then evaluations are continued with the Q-factor which is the key indicator of broad Smith-chart impedance coverage. EM Co-simulations show better performance in terms of estimating measurement results compared to schematic simulations. Measurement results demonstrate Q-factor of the parasitic injected designs have a higher Q-factor than those without parasitic injection. This reveals that proper circuit synthesis is more important than reducing the number of components in series branches to obtain a higher Q-factor. However, reducing components in series branches still contributes to obtaining a higher Q-factor. The highest Q-factor is obtained on the modified design with parasitic injection with 70. The nonlinear behavior of impedance matching networks is also evaluated by determining output 1-dB compression points. A measurement set-up is constructed by using a preamplifier and the output of the amplifier is connected to matching networks. Measured results reveal that initial designs have almost the same 1-dB output compression point around 24.5±0.2 dBm, where the modified design with parasitic injection reaches the highest with 25.8 dBm and without parasitic injection has the lowest value with 23.5 dBm. Finally, Smith-chart impedance coverage of each matching network is examined with simulations and measurements. In order to measure the impedance coverage, an algorithm is proposed that measures the input impedance of matching networks under every possible combination of varactor diode reverse bias voltage from 0 V to 8 V with 0.2 V steps. Schematic simulations are in better agreement with the measurement results. Q-factor measurement results are aligned with the Smith-chart coverage which verifies the design concerns. The highest impedance coverage is obtained with the parasitic injected modified design which provides at least 50 % impedance coverage between 0.9 GHz and 2.2 GHz. Comparison with similar works reveals that the designed reconfigurable impedance matching network has superior performance in terms of frequency tunability, nonlinearity, insertion loss, and noise figure. This improvement is achieved by using a proper circuit model for enhanced tuning elements and detailed circuit synthesis. The modified design with parasitic injection which has the best performance among the other designs is used in antenna mismatch compensation. A simple patch antenna matched to 50Ω is simulated and measured in free space. Then, in proximity to the human head, the change in the input impedance of the antenna is observed with simulations and measurements. When the antenna is close to the human head, the input impedance is no longer matched to 50Ω. In that case, the transceiver has to consume more power to ensure communication. However, with the help of the designed matching network, input impedance change is compensated by tuning the input impedance to 50Ω for efficient power transmission. Moreover, a center frequency tuning feature is provided to the antenna with the help of the designed reconfigurable impedance matching network.