Gavin T. Watkins

A 60% PAE WCDMA Handset Transmitter Amplifier

This paper reports the design of a class-E envelope elimination and restoration (EER) based amplifier for a wideband code division multiple access handset application that attains 60% power-added efficiency at peak power output. Emphasis is placed on the envelope modulator that employs a novel split-frequency approach in order to attain an efficiency of 80% for this part of the system. In contrast to standard EER systems, the carrier is not amplitude limited, but rather predistorted to maintain both linearity and power efficiency. Performance in terms of efficiency, spectral output, and error vector magnitude is reported.

Cognitive antenna selection relay for green heterogeneous healthcare networks

This article presents the latest progress on green healthcare research in heterogeneous networks (HetNet), where devices are capable of switching between multiple radio access technologies (RAT) and each RAT operates on a different frequency channel. After outlining the design features and challenges of a medical service paradigm for green cognitive medical body area networks (MBAN), measurements are made to investigate the turn-on characteristics of a power amplifier (PA) in terms of excess energy consumption and turn-on delay. We then present a multi-mode PA architecture for an MBAN relay system. An energy-efficient PA switch/stay mechanism is also presented. This allows the proposed PA architecture to operate given a transmission outage probability and transmission delay constraint. Antenna selection between two heterogeneous RATs is exploited to improve the transmission reliability. Both measurement and numerical results are provided to corroborate the performance of the proposed architecture and its switch/stay mechanism.

Design method for harmonically-tuned, dynamic load-modulated power amplifiers

In this paper, an alternative technique for the design of harmonically-tuned, dynamic load-modulated power amplifiers is proposed, investigated and compared against traditional load-pull in simulations. The method exploits the flexibility of “continuous” modes of amplifier operation while it optimises the harmonic terminations for back-off performance. The large signal transistor model of a 10W GaN device (CGH40010) is used to simulate performances at two frequencies of 0.9GHz and 2GHz. The simulated efficiencies and intrinsic waveforms show that the proposed method represents a more comprehensive design path for harmonically tuned dynamic load-modulated PAs. Moreover, it can lead to optimised performance for the amplification of modulated signals.

Improved Reactance-Compensation Technique for the Design of Wideband Suboptimum Class-E Power Amplifiers

A reactance-compensation technique has been introduced recently for the design of wideband class-E power amplifiers (PAs). With this technique, the load resistance can be transformed to an optimal complex drain impedance in a broad frequency band. One potential problem of this technique is that an additional network is often needed to transform 50 Ω to the optimum load resistance, which is typically at a lower value. This paper proposes a method to improve the reactance compensation technique. By using the proposed method, the required low-value load resistance is up to four times the original value. This is achieved not by adding more components, but by simply changing the order and values of them. With the proposed method, the additional resistance matching network is no longer needed in many cases, or can be significantly simplified in other cases. To validate the theory, two broadband class-E amplifiers were designed using the original technique and the proposed method, respectively. The performances of the two amplifiers are compared. By using similar complexity of matching networks, the PA designed using the original method can achieve better than 70% power-added efficiency for a fractional bandwidth of 42%. The PA designed using the proposed method can achieve above 70% efficiency for a bandwidth of 51%. The proposed method can be used in the design of many other types of amplifiers.

Bandwidth reduction in dynamic load-modulated power amplifiers: control and RF signal expansion, efficiency and linearity trade-offs

This paper presents an experimental investigation of the effect of bandwidth reduction for baseband control signals used in dynamic load-modulated (DLM) amplifiers. Using a low-power reactively DLM architecture operating at 2.35GHz the effect of control signal filtering on efficiency, output power and adjacent channel power ratio (ACPR) of the system is shown for various signal bandwidths. The method is applied on WCDMA and LTE signals showing a degradation of 2-4 percentage points (pp) in efficiency, 0.5dB lower output power and similar ACPR performance for reducing the bandwidth of the control signal by 70%.

Energy-efficient heterogeneous antenna selection relaying in wireless body area networks

We present a multi-mode power amplifier (PA) set with an antenna selection mechanism in a body area relaying network for healthcare applications. The PA set of the relay node is designed to work in two modes where the RF switch utilized for antenna selection is located between the highly efficient narrowband PAs and a low-efficiency wideband PA supporting two heterogeneous networks. Our study shows that, unlike in the wideband PA mode, where antenna selection is always desirable for transmission reliability, antenna selection is not always preferred in the narrowband PA mode since in this mode, the PA needs to turn on and off constantly. Consequently, extra switch delay and energy consumption occur. To balance the tradeoff between the relay nodes energy efficiency and transmission reliability, both the switching probability between two modes and bounds on the probability of staying on the current antenna are analyzed. Numerical results are provided to corroborate the improved performance of the proposed architecture and the switch/stay mechanism.

Flexible linearity profile low noise feedforward amplifiers for improving channel capacity

This paper describes a highly linear low noise amplifier (LNA) exhibiting a flexible linearity profile intended as a way to meet future flexible mobile communications standards. This has the potential of inclusion in ad-hoc software defined radio (SDR) networks where a frequency band and modulation scheme can be chosen based on the characteristics of the frequency spectrum and user requirements. The linearity profile of the LNA is tailored to cancel distortion generated as a function of the received power profile. This improves the receivers tolerance to unwanted signals present in the frequency band of interest and reduces the requirement of coding and equalization to compensate for receiver limitations whilst increasing the available capacity. The flexible linearity profile is achieved by the use of a feedforward amplifier structure which has the capability of large distortion suppression determined by the gain and phase balances within its two loops. To ensure maximum distortion cancellation in the band, adaptive cancellation is used to maximize the cancellation of harmful distortion.

Load Modulation of Harmonically Tuned Amplifiers and Application to Outphasing Systems

Modulation of load impedance is an effective way to maintain efficient power amplifier (PA) operation over high dynamic range modulated signals. For high efficiency, a load modulation approach can be applied to inherently efficient classes of PAs such as those with harmonic tuning: Class J, Class F, and Inverse Class F. This paper presents an analysis of harmonically tuned (HT) amplifiers operating under load modulation conditions, deriving the optimal loading trajectories for these multiple classes of operation. Because these load trajectories are complex, it is then shown—through a series of analysis, simulations, and measurements—that HT amplifiers are better suited for outphasing systems, than conventional amplifiers such as Class B. A design methodology is proposed and validated through design and measurement of a 900-MHz outphasing system, comprising of two Gallium Nitride Class J branch PAs, delivering 44.6 dBm with 75% PAE at saturation, while mantaining PAE above 60% over a 7-dB output power back-off.

Asymmetrical outphasing: Exploiting conjugate continuous modes of operation

In conventional techniques for efficiency enhancement of outphasing systems, identical power amplifiers are employed for each branch while the output combining network is optimised using reactive compensating elements. In this work a methodology is presented for asymmetrical branch amplifier design based on the conjugate load modulation trajectories seen by each PA branch. Closed form equations and simulations are presented, demonstrating that conjugate modes of operation PAs can be used for the upper and lower amplifier branches in outphasing systems to achieve ideal maximum efficiency over the outphasing backoff range. As a proof of concept, Class B and Class J PAs were built using the Cree CGH40010 GaN HEMT. The Class J PA exhibited over 15% improvement in drain efficiency at 4dB back-off compared to the Class B PA when both were measured with the lower branch outphasing load trajectory.

A single input transformer-less push-pull microwave power amplifier

A transformer-less push-pull RF power amplifier (PA) is described based on the totem-pole configuration. This involves stacking two transistors one on top of another and driving them in anti-phase. A variable attenuator and delay line are included so that the gain and phase relationship of the signals driving the two transistors can be accurately manipulated. It is shown, that as the gain and phase are swept, an optimum exists where efficiency is maximum and the second harmonic distortion (HD2) suppressed. These are the metrics used to define correct push-pull operation. At the optimum setting, a practical amplifier operating at 680 MHz achieved an efficiency of 52% at 23.6 dBm output power (POUT), with 48 dB HD2 suppression.