This application claims priority from South African provisional patent application number 2016/02559 filed on 15 Apr. 2016, which is incorporated by reference herein.
In wireless communication systems, power amplifiers (PAs) may be used in the final stage of a transmitter to increase power in a transmitted signal. The PAs typically include an active device such as a transistor and convert direct current (DC) power into radio-frequency (RF) power while being driven by a RF input signal. The portion of DC power that is not converted into RF power is dissipated as heat and constitutes a loss of power. The power conversion mechanism that takes place in a PA is described by the power conversion efficiency (PE) and the power added efficiency (PAE).
The power conversion efficiency is typically formulated as the percentage of the DC power that is converted into useful RF output signal power, while the power added efficiency is typically formulated as the percentage of DC power that is converted into useful RF output signal power while taking into account the effect of the RF input (or drive) signal.
It is often the case in wireless communication systems that the PA is the primary consumer of DC power. The power conversion efficiency is therefore an important performance parameter of the PA. A higher power conversion efficiency may for example lead to longer battery lifetime, simpler thermal management requirements and reduced form factor of the PA, amongst others.
It has been shown that the power conversion efficiency can be maximised by ensuring that the DC power is equal to the power ultimately delivered to the load at the fundamental frequency.
This occurs when the sum of power dissipated in the active device and the power delivered to the load at harmonic frequencies is zero. It is known that two necessary and sufficient conditions may fulfil these requirements: firstly by ensuring that the drain voltage and current of the active device do not overlap and secondly by terminating the drain of the active device with appropriate impedances at harmonic frequencies.
These two necessary and sufficient conditions can be fulfilled by ensuring that the drain voltage of the active device is a half sinusoidal waveform and the current waveform is square. PAs typically include a load network which is designed to aid fulfilment of these conditions. An inverse Class-F PA can fulfil the necessary and sufficient conditions to obtain the maximum theoretical power conversion efficiency if the load network presents the following impedances to the drain of the active device:
where Zn is the impedance presented by the load network to the drain terminal of the active device, VDD is the drain voltage, I0 is the drain current at the fundamental frequency, Ropt is the theoretical optimal resistance and n is the order of harmonic.
However, there are some load network designs which meet the abovementioned necessary and sufficient conditions to some degree. As a result, the efficiency of existing PAs vary somewhat.
Accordingly, there is scope for improvement.
In accordance with the invention there is provided a power amplifier comprising:
an input for receiving a signal to be amplified, the signal being associated with a fundamental frequency;
an amplifier circuit including an active device for amplifying the input signal;
an output for providing the amplified signal to a load; and,
a load network electrically interposed between the amplifier circuit and the output and including:
A further feature provides for the fundamental frequency matching circuitry to include a first portion and a second portion, the parallel transmission line arrangement being interposed between the first and second portions.
The power amplifier including the one-eighth wavelength stubs, each respectively extending from a first junction which is electrically coupled to the fundamental frequency matching circuitry and to the output.
The parallel transmission line arrangement operatively presents, at the first junction, an open-circuit condition at the fundamental frequency, a short-circuit condition at a second harmonic of the fundamental frequency and an open-circuit condition at a third harmonic of the fundamental frequency.
A further feature provides for the fundamental frequency matching circuitry to include a first series-connected transmission line having a first electrical length at the fundamental frequency and being electrically interposed between the amplifier circuit and the first junction; a second series-connected transmission line having a second electrical length at the fundamental frequency and being electrically interposed between the first junction and the output; and, at the fundamental frequency, a one-twelfth wavelength open-circuit stub electrically interposed between the second series-connected transmission line and the output.
An additional feature provides for the power amplifier to be an inverse Class F power amplifier.
The invention extends to a load network for an inverse Class-F power amplifier, the load network being arranged to be electrically interposed between an amplifier circuit and an output of the inverse Class-F power amplifier and comprising:
fundamental frequency matching circuitry which presents an optimal resistance at the fundamental frequency; and,
a parallel transmission line arrangement, the parallel transmission line arrangement including, at the fundamental frequency, a one-eighth wavelength short-circuited stub and, at the fundamental frequency, a one-eighth wavelength open-circuit stub, each of which extends respectively from a first junction which is electrically coupled to the fundamental frequency matching circuitry and which is arranged to be electrically coupled to the output,
wherein the fundamental frequency matching circuitry is arranged to cooperate with the parallel transmission line arrangement such that the load network operatively presents:
The parallel transmission line arrangement operatively presents, at the first junction, an open-circuit condition at the fundamental frequency, a short-circuit condition at a second harmonic of the fundamental frequency and an open-circuit condition at a third harmonic of the fundamental frequency.
A further feature provides for the fundamental frequency matching circuitry to include a first series-connected transmission line having a first electrical length at the fundamental frequency and being configured to be electrically connected to the amplifier circuit and which is connected to the first junction; a second series-connected transmission line having a second electrical length at the fundamental frequency and being electrically connected to the first junction and being configured to be electrically connected to the output; and, at the fundamental frequency, a one-twelfth wavelength open-circuit stub electrically interposed between the second series-connected transmission line and the output.
An embodiment of the invention will now be described, by way of example only, with reference to the accompanying drawings.
In the drawings:
The PA (100) may include an input matching network (104) which is arranged to match the input impedance of the PA (100) to the source impedance. The input matching network (104) may be provided by a suitable arrangement of transmission lines. One exemplary arrangement includes a series stub followed by an open-circuit shunt stub, although other arrangements may also be possible.
The PA (100) includes an amplifier circuit (106). The amplifier circuit (106) includes an active device (108) which is arranged to amplify the input signal. The active device (108) may for example be a transistor, such as a gallium nitride (GaN) high-electron-mobility transistor (HEMT). The active device (108) is electrically coupled to a direct-current (DC) supply, the input (102) (via the input matching network (104)) and an output (110) of the PA (100), which is arranged to provide the amplified signal to a load (112), such as an antenna.
The PA (100) includes a load network (114) which is electrically interposed between the amplifier circuit (106) and the output (110). The load network (114), which is illustrated in greater detail in
The fundamental frequency matching circuitry (120) is arranged to present the optimal resistance (Ropt) at the fundamental frequency (f0). As will be explained in greater detail below, the fundamental frequency matching circuitry (120) is further arranged to cooperate with the parallel transmission line arrangement (130) such that the load network (114) operatively presents, to a drain of the active device, an optimal resistance at the fundamental frequency; an open-circuit at a second harmonic of the fundamental frequency; and, a short-circuit at a third harmonic of the fundamental frequency.
In one implementation, the fundamental frequency matching circuitry (120) may include a first series-connected transmission line (122) which has a first electrical length (θ1) at the fundamental frequency (f0). The first series-connected transmission line (122) is electrically interposed between the amplifier circuit (106) and a first junction (124). The fundamental frequency matching circuitry (120) may also include a second series-connected transmission line (126) having a second electrical length (θ2) at the fundamental frequency (f0). The second series-connected transmission line (126) is electrically interposed between the first junction (124) and the output (110). The fundamental frequency matching circuitry (120) may further include, at the fundamental frequency (f0), a one-twelfth wavelength
open-circuit stub (128) which is electrically interposed between the second series-connected transmission line (126) and the output (110). Each of the transmission lines has an associated impedance.
The parallel transmission line arrangement (130) includes, at the fundamental frequency (f0), a one-eighth wavelength
short-circuited stub (132) and, at the fundamental frequency (f0), a one-eighth wavelength
open-circuit stub (134). Each of the open-circuit and short-circuited stubs (132, 134) extend respectively from the first junction (124) which is electrically coupled to the fundamental frequency matching circuitry (120), in-between the first and second series-connected transmission lines (122, 126), and in turn to the output (110). The open-circuit and short-circuited stubs (132, 134) of the parallel transmission line arrangement (130) have associated impedances.
In operation and at the fundamental frequency (f0), the parallel transmission line arrangement (130) presents an open-circuit condition at the first junction (124). The effect of this is to reduce the load network (114) to an equivalent circuit (140), illustrated in
open-circuit stub (144). The series-connected transmission line (142) has an electrical length equivalent to that of the sum of the first and second series-connected transmission lines (122, 126) (θ1+θ2) of the fundamental frequency matching circuitry (120). The equivalent circuit (140) includes the shunt capacitor (CDS) and series inductor (Lout), a source (146) and a load (RL).
The equivalent circuit (140) can be reduced further to the circuit (150) illustrated in
open-circuit stub (144) is modelled as a shunt capacitor (C) whose value is given by:
and in which the series-connected transmission line (142) is modelled as a series inductor (L) with a value given by:
In order for the load resistance (RL) to match the optimum resistance (Ropt) at the fundamental frequency, the capacitive and inductive components of the resulting low pass π-type matching network (150) can be calculated using the following equations:
where the source quality factor (QS) and the load quality factor (QL) are given by the following two equations:
The above two equations assume that the load resistance (RL) is greater than the optimum resistance (Ropt). However, this assumption can be disregarded if the source quality factor (QS) is known.
In operation and at the second harmonic of the fundamental frequency (2f0), the short-circuited stub (132) and the open-circuit stub (134) have a one-quarter wavelength
and the parallel transmission line arrangement (130) therefore presents a short-circuit condition at the first junction (124) at this harmonic. This short-circuit condition results in the equivalent load network (160) which is illustrated in
In order to achieve the required open-circuit condition at the second harmonic (2f0) at the drain of the active device (108), the admittance (Yin) presented by the equivalent load network (160) must be equal to zero. That is:
Now considering operation at the third harmonic of the fundamental frequency (3f0), the short-circuited stub (132) and the open-circuit stub (134) behave as they would at the fundamental frequency (f0) and the parallel transmission line arrangement (130) therefore presents an open-circuit condition at the first junction (124). Furthermore, the open-circuit stub (128) of the fundamental frequency matching circuitry (120) has a one-quarter wavelength
electrical length and thus presents a short circuit condition at a second junction (168) of the load network (114). This results in an equivalent load network (170) for the third harmonic (3f0) which is illustrated in
The equivalent load network (170) at the third harmonic (3f0) includes the shunt capacitor (CDS) and series inductor (Lout) as well as a source (172). The equivalent load network (170) at the third harmonic (3f0) also includes a series-connected transmission line (174) which has an electrical length (3(θ1+θ2)) which is equivalent to thrice that of the first and second series-connected transmission lines (122, 126) in combination.
In order to achieve the required short-circuit condition at the drain of the active device (108) at the third harmonic (3f0), the impedance (Zin) presented by the equivalent load network (170) illustrated in
j3ω0Lout+Z1tan 3(θ1+θ2)=0 (9)
For this condition to be true, the combined electrical length (3(θ1+θ2)) of the series-connected transmission line (174) must be equal to 180° at the third harmonic (3f0) and thus equal to 60° at the fundamental frequency (f0) when parasitic inductance (Lout) is equal to 0 H.
Manipulating equations (3) (8) and (9) above result in a set of two simultaneous equations with two unknowns, θ1 and θ2:
Solving the above two equations (10) and (11), using the known constants relevant to the particular PA, will produce the required parameters for the load network (114). In what follows, an exemplary inverse Class-F PA is described and the above two equations are solved to arrive at the relevant parameters for the load network for the PA.
Design of an Exemplary Load Network
The proposed load network was designed to operate at a fundamental frequency (f0) of 2.2 GHz. The optimum resistance (Ropt) value for a Class-F1 PA is determined using the load line technique, that is, by determining the ratio of the drain voltage (VDD) and current (I0) at the fundamental frequency (f0). For the exemplary PA, an optimum resistance (Ropt) value of 90.428 Ω was obtained using equation (1). It should however be appreciated that instead of using the load line technique, the load pull technique may also be used to determine the optimum impedance (Zopt) to maximize power conversion efficiency at the fundamental frequency (f0). In such a case, the above described load network (114) can be modified to accommodate an additional output matching network.
In this exemplary PA, a Cree CGH40010F GaN active device was provided which has a drain capacitance (CDS) and lead inductor (Lout) equal to 1.3 pF and 0.653 nH respectively. The values of the capacitor (C) and inductor (L) of the low pass π-type equivalent load matching network (150) illustrated in
Solving the set of equations in (10) and (11), results in the electrical length (θ1) of the first series-connected transmission line (122) being equal to 3.961°and the electrical length (θ2) of the second series-connected transmission line (126) being equal to 49.375°.
Initially, an assumption was made that the two series-connected transmission lines (122, 126) would have the same impedance (Z1) in order to simplify the design of the load network. Based on this, the impedance (Z1) of each of the two series-connected transmission lines (122, 126) of the fundamental frequency matching circuitry (120) was determined, using equation (3), to be 70.22 Ω respectively. However, an optimization process, described below, was then performed to take into account this assumption.
The characteristic impedance (Z0) of the open-circuit and short-circuited stubs (132, 134) of the parallel transmission line arrangement (130) was selected to be 50 Ω.
Suitable electronic design automation software, such as Advanced Design System™, was used to perform simulations on the designed load network. The simulated input reflection coefficient of the exemplary load network, with the parameters determined above, is illustrated in
The initial simulation results show that the proposed load network with the parameters described above achieves the required impedance termination for an inverse Class-F PA. However, the optimum resistance (Ropt) of 90.428 Ω is not well matched to the 50 Ω load resistance (RL). This is due to the assumption described above, that is, that the two series-connected transmission lines (122, 126) would have the same characteristic impedance (Z1). This assumption was made to simplify the design of the load network. In order to account for the assumption, the load network was optimized using electronic design automation software. Results of the optimization are presented in the table below.
The optimized simulated input reflection coefficient of the exemplary load network, with the parameters described in the table above, is illustrated in
The optimized simulated input reflection coefficient loss illustrated in
The load network described herein provides a parallel transmission line arrangement (130) which includes an open-circuit one-eighth wavelength transmission line in parallel with a short circuited one-eighth wavelength transmission line. In a comparison between a conventional quarter wave impedance transformer and the above described parallel transmission line arrangement (130), it was found that the rejection band of the parallel combination of an open and a short-circuited one-eighth wavelength stub is twice that of the quarter wave impedance transformer. The physical length of the one-eighth wavelength stub is half that of the quarter wave impedance transformer. Thus a lower electrical series resistance (ESR) results in turn leading to a better short circuit condition. The length of the parallel transmission line arrangement (130) described herein may also reduce the physical size of the PA.
A one-eighth wavelength transmission line with a short circuit load is equivalent to an ideal capacitor, that is, the input impedance the transmission line (Zin) is equal to 0−j(ωC)−1 Ω. Similarly, a one-eighth wavelength line with an open circuit load is equivalent to an ideal inductor, that is, the input impedance the transmission line (Zin) is equal to 0+j(ωL) Ω. The arrangement of the one-eighth wavelength transmission lines described herein is such that the equivalent circuit is a parallel resonant circuit. The characteristics of a parallel resonant circuit are that it presents a very high impedance at its input port at resonance and very low impedances at its input port at signal frequencies below and above the resonant frequency. Further, the actual resistance of the copper of the actual transmission line sections may be very low, which results in a very high Q, which in turn results is a very narrow bandwidth, thereby ensuring that only the fundamental component of the drain signal reaches the load resistance and all other even harmonics are prevented from getting reaching the load.
Accordingly, the parallel transmission line arrangement (130) including the one-eighth wavelength short circuit transmission line and the one-eighth wavelength open-circuit transmission line described herein forms a very high Q parallel resonant circuit with a very high input impedance at the fundamental frequency and all odd harmonics.
At all even harmonics the parallel transmission line arrangement (130) forms a short-circuit path to ground which prevents these even harmonic components from reaching the load. Accordingly, the inverse Class-F PA described herein may better meet the necessary and sufficient conditions for improved efficiency over existing PAs.
When compared to one of the best performing commercially available Class-F−1 PA's, the Gebrennikov PA, the practically implemented PA (300) of the present disclosure performs significantly better. The practically implemented PA of the present disclosure has a PE of 95% (as shown in Table 2), whereas the Gebrennikov PA, when implemented, has a PE of 76% using exactly the same active device. Similarly, the PAE attained by the practically implemented PA (300) is 82% which is 10% higher than the PAE achieved by the Gebrennikov PA.
As mentioned above, the load network described in the foregoing was designed using the load line technique to determine the optimum resistance for maximum power conversion efficiency at the fundamental frequency. However, the load pull technique can also be used to determine the optimum impedance for maximum power conversion efficiency at the fundamental frequency. In such a case the proposed load network would be modified to accommodate an additional output matching network.
The load network described herein may be implemented in any suitable inverse Class-F PA. In some cases, the load network may be implemented as a part of an inverse Class-F PA used in a wireless communication system. Exemplary wireless communication systems include terrestrial systems, such as cellular communication networks and the like, or satellite-based systems. It should be appreciated that the inverse Class-F PA herein described may find application in other technology sectors.
The term “load network” used herein may also be referred to as a wave shaping network or an output matching network. The term “load network” may refer to any suitable arrangement of components whose function is to ensure that the drain of an active device of an inverse Class-F PA is terminated with the optimum impedance at the fundamental frequency, an open-circuit at even harmonic frequencies and a short circuit at odd harmonic frequencies. The load network should also absorb the parasitic elements of the active device to minimize their effect on the drain power conversion efficiency.
“Transmission lines” referred to herein include any suitable structure designed to carry alternating current with a frequency high enough that their wave nature must be taken into account. The transmission lines referred to herein may be provided by a coaxial cable, stripline, microstrip, waveguide, dielectric waveguide or the like.
The term “stub” as used herein refers to a length of transmission line which is connected at one end only, the free end being either left open-circuit or short-circuited.
The foregoing description has been presented for the purpose of illustration; it is not intended to be exhaustive or to limit the invention to the precise forms disclosed. Persons skilled in the relevant art can appreciate that many modifications and variations are possible in light of the above disclosure.
Some portions of this description describe the embodiments of the invention in terms of algorithms and symbolic representations of operations on information. These algorithmic descriptions and representations are commonly used by those skilled in the relevant art to convey the substance of their work effectively to others skilled in the art.
The language used in the specification has been principally selected for readability and instructional purposes, and it may not have been selected to delineate or circumscribe the inventive subject matter. It is therefore intended that the scope of the invention be limited not by this detailed description, but rather by any claims that issue on an application based hereon. Accordingly, the disclosure of the embodiments of the invention is intended to be illustrative, but not limiting, of the scope of the invention, which is set forth in the following claims.
Finally, throughout the specification unless the contents requires otherwise the word ‘comprise’ or variations such as ‘comprises’ or ‘comprising’ will be understood to imply the inclusion of a stated integer or group of integers but not the exclusion of any other integer or group of integers.
Number | Date | Country | Kind |
---|---|---|---|
2016/02559 | Apr 2016 | ZA | national |
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/IB2017/052131 | 4/13/2017 | WO | 00 |