Patent Grant
9564935
This invention relates to a power amplifier and a transmitter for use in a base station and with terminals of a mobile wireless communication system (e.g. micro-wave mobile phones), a satellite communication system, and a broadcasting system.
More particularly, the invention relates to a composite power amplifier for linear amplification of an RF modulated signal, e.g. a multi-carrier signal having a large peak envelope power as compared with its average output power. The inventive composite transmitter is simpler in structure than conventional ones, yet it has an improved linearity and power efficiency characteristics.
The present invention pertains to an improvement of a composite power amplifier such as a Chireix power amplifier (See Non-patent Document 1) and a Doherty power amplifier (Non-patent document 2), the improvement directed to remove a disadvantageous narrow band characteristic of such composite power amplifier. The Chireix power amplifier was invented by Chireix in 1935, and the Doherty power amplifier was invented by Doherty in 1936. Both of them have been used as AM broadcasting transmitters. Before semiconductors were invented, their power amplifying elements were vacuum tubes. However, after semiconductor amplifiers were invented, they were widely used as final-stage amplifiers, especially in mobile, satellite, and broadcasting systems. The Chireix amplifiers are often called LINC power amplifiers (Non-patent Reference 3).
In the Chireix amplifiers and the Doherty amplifier (hereinafter referred to as composite amplifiers), the ratio between an output voltage and an output current, known as equivalent load impedance, is varied in accordance with the magnitude of an input signal, thereby increasing the power efficiency of an amplifier. This makes their amplification behaviors extremely complex, so that it is difficult to maintain a high power efficiency and a linearity over a wide frequency band.
An operational principle and a power efficiency characteristic of an prior art composite power amplifier will now be given below with reference to the Chireix amplifier first and then to the Doherty amplifier.
As shown in
(1) a signal component separating device 190 which, upon receipt of an in-phase component I(t) of a baseband signal (the in-phase component I(t) hereinafter referred to as in-phase signal I(t)) and an orthogonal component Q(t) of the baseband signal (the orthogonal component hereinafter referred to as orthogonal signal Q(t)), generates
an RF modulated signal A (hereinafter referred to a main signal A), obtained by orthogonal modulation of I(t) and Q(t) (Eq. 1 below),
a signal B orthogonal to the main signal A (the orthogonal signal B hereinafter referred to as efficiency improving signal EIS B), and
a first composite signal S1 in the form of a vectorial sum of the main signal A and the EIS B, and
a second composite signal S2 in the form of a vectorial difference between the main signal A and the EIS B;
(2) power amplifiers 150 and 151 for power amplifying the respective composite signals S1 and S2; and
(3) a Chireix power combiner 140 for combining the outputs of the power amplifiers 150 and 151 to provide the output signal S0 of the Chireix transmitter 101.
The Chireix network 140 consists of two impedance inverters (or ¼-wavelength lines) 160 and 161 each adapted to interface its input terminals and its output terminal; and two reactance elements 170 and 171 each adapted to interface its respective input terminals and the ground. The reactances of the two reactance elements have the same absolute value, but have opposite signs. It is noted that the term “efficiency improving signal (EIS)” is not a common technical term but is a term named by the present inventor to connote an implication of the signal that it enhances the power efficiencies of the power amplifiers. This term will be used in this context in an exposition below of the prior art Doherty transmitter and in the description of the present invention as well.
The main signal A is a high-frequency modulated signal (or RF modulated signal) obtained by an orthogonal modulation of an in-phase component of a baseband signal (hereinafter referred to as in-phase baseband signal) I(t) and an orthogonal component of the baseband signal (hereinafter referred to as orthogonal baseband signal) Q(t) with a carrier angular frequency ω0. Thus, the main signal A is given by the following equation.
A(t)=I(t)cos(ω0t)+Q(t)sin(ω0t) (Eq. 1)
In terms of an envelope signal a(t) and a phase modulation signal φ(t), the main signal A and its vectorial form a can be written in the following forms.
A(t)=a(t)cos {ω0t+φ(t)} (Eq. 2)
a=a(t)exp{jω(t)} (Eq. 3)
An EIS B, which is orthogonal to the main signal A, satisfies a condition that the envelope of the composite signal of the main signal A and the EIS B equals the peak envelope level C (which is often referred to as peak level C or simply C) of the main signal A. From this condition, the EIS B is given by the following equation.
B(t)=−b(t)sin {ω0t+φ(t)} (Eq. 4)
where b(t)=√{square root over (C2−a(t)2)} and C is the peak level of the main signal.
The power amplifiers 150 receives the first input signal S1 which is a vectorial sum of the EIS B (for which a vector representation is jb) and the main signal A, while the power amplifier 151 receives the second input signal S2 which is a vectorial difference between the main signal A and the EIS B, as given by the following Equations.
S1(t)=A(t)+B(t) (Eq. 5)
S2(t)=A(t)−B(t) (Eq. 6)
Plugging Eq. 2 and Eq. 4 in Eq. 5 and Eq. 6, respectively, S1(t) and S2(t) turn out to be
S1(t)=C cos [ω0t+φ(t)+cos−1{a(t)/C}] (Eq. 7)
S2(t)=C cos [ω0t+φ(t)−cos−1{a(t)/C}] (Eq8)
Eq. 7 and Eq. 8 show respectively that the envelope levels of the first and the second composite signals S1 and S2, respectively, are constant (equal to the peak level C of the main signal) and that they are either advanced or delayed in phase relative to the main signal by cos−1{a(t)/C}.
Since in the Chireix transmitter 101 the envelope levels of the input signals fed to the power amplifiers 150 and 151 are always equalized to the peak level C of the main signal by adding the EIS B to the main signal A, the power amplifiers 150 and 151 always operate at their maximum power efficiencies. The EISs B fed to the input ends of the power amplifiers 150 and 151 are cancelled out at the output end of the Chireix power combiner 140, resulting in an output signal S0 given by the following Eq. 9.
S0(t)=g{S1(t)+S2(t)}/√{square root over (2)}=√{square root over (2)}gA(t) (Eq. 9)
where g is the voltage gain of the power amplifiers 150 and 151.
Focusing attention to the power amplifier 150, it appears that, because of the power amplifier 151 in operation, an equivalent output impedance (i.e. the ratio between the output voltage and the output current) of the power amplifier 150 turns out to be a non-real number. In other words, it appears that an element of reactance Xc is connected between the output end of the power amplifier 150 and the ground. In order to achieve an impedance matching with the power amplifier 151, it suffices to connect an reactance element 170 of reactance −Xc between the output end of the power amplifier 150 and the ground. In this case, however, the magnitude of Xc changes with the envelope level a(t) of the main signal A, so that the impedance matching can be achieved only when the main signal A has certain particular magnitudes. Referring to
Next, a Doherty transmitter 103 will now be discussed, which is one of variable-load impedance type composite transmitters equally well known as the Chireix transmitter. Referring to
(1) an orthogonal modulator 90 which, upon receipt of an in-phase baseband signal I(t) and an orthogonal baseband signal Q(t), outputs a main signal A obtained by orthogonal modulation of the two input signals;
(2) a power amplifier 152 for power amplifying the main signal A fed thereto;
(3) a power amplifier 153 for power amplifying the main signal A fed thereto after the main signal A is delayed by means of a ¼-wavelength line 163; and
(4) a Doherty power combiner 142 for combining the outputs of the power amplifiers 152 and 153 via an impedance inverter 162 to thereby provide a transmission output signal S0.
In the Doherty transmitter 103, the power amplifier 152 is called a carrier amplifier (CA), which is in practice either B- or AB-class amplifier. The power amplifier 153, which is also called peaking amplifier (PA), is a C-class CA. Each of the CA and the PA receives a bifurcated and power-amplified main signal A. The Doherty transmitter 103 can be represented by an ideal current source model as shown in
Operations of the Doherty transmitter 103 may be discussed in two separate domains, one in a small power domain (where the envelope level is not more than ½ of the peak vale C of the signal) and in a large power domain (where the envelope level of the main signal exceeds ½ of the peak level.) Referring to curve c50 shown in
As the voltage of the main signal exceeds one half the peak envelope level C, the PA begins to operate, causing a further current to be supplied from the PA to the Doherty transmitter 103, thereby reducing its apparent load impedance. As the CA remains saturated at a saturation point, maintaining a constant voltage, the CA can be regarded as a constant-voltage power source that operates at its maximum power efficiency. When outputting a peak envelope power (PEP), the CA and the PA can see a load of 50Ω, each providing its power as much as ½ of the maximum output power, and then the theoretical PEP efficiency of the CA is 78.5% if the CA is a B-class amplifier (Non-patent Document 6).
It has been disclosed in a literature (Non-patent Document 6) that when the Doherty transmitter 103 shown in
In recent years, it has been proposed in Non-patent Document 6 (
A merit of the Doherty transmitter 104 lies in the fact that the input power fed to the power amplifiers 152 and 153 can be reduced as compared with the Doherty transmitter 103, which helps increase the power add efficiency of the Doherty transmitter 104, and that the same B- or AB-class amplifiers can be used as the two power amplifiers 152 and 153, which permits reduction of its manufacturing cost and extension of operable frequency band, as disclosed in the Patent Document 1.
The EIS to be added to the main signal A of the Doherty transmitter 104 will be referred to as EIS A1, as in the Chireix transmitter 101. By writing the main signal A in the form of Eq. 1 and Eq. 2 as in the Chireix transmitter 101, the EIS A1 is given by
The first composite signal S1(t) fed to the power amplifier 152 is a vectorial sum of the main signal A and the EIS A1, while the second composite signal S2(t) fed to the power amplifier 153 is a vectorial difference between the main signal A and the EIS A1, which are give by the following equations, respectively
Patent Document 1 proposes to replace, by an alternative Doherty transmitter 105 (not shown), a circuit of the Dohery transmitter 104 that generates a first and a second composite signals S1 and S2, respectively, from a main signal. The Doherty transmitter 105 generates the second composite signals S2 by passing the main signal A through a nonlinear circuit (nonlinear emulator 181), and then generates the first composite signal S1 by passing S2 through a cross combining filter 182 and subtracting the resultant signal from the main signal A.
In a conventional Doherty transmitter 103, the input vs output voltage characteristic of the CA for inputted main signal is not constant in a large-power domain, while the input vs output voltage characteristic of the PA lacks linearity in a small-power domain, as shown in
In any of conventional composite transmitters discussed above, a usable frequency bandwidth of the EIS is limited by a limitative operational speed of a digital signal processing circuit used. Furthermore, in actuality the first nor the second power amplifiers can never be a perfect linear amplifier, outputting nonlinear distorted components even outside the limited frequency band of the EIS.
This phenomenon is called “spectral regrowth”, which will be now described with reference to Fig. (a)-(b) for a case where the first and the second power amplifiers are of AB-class (having a gate bias voltage 0.2 times that of an A-class amplifier), and a main signal is an standard RF modulated signal (or 4-wave QPSK signal, as defined and used in the first embodiment below).
Patent Document 1:
As discussed above, prior art composite amplifiers are so-called variable-impedance amplifiers. That is, the output impedances of its two power amplifiers change with the level of the main signal. Since the structure of a power combiner for combining two outputs of two power amplifiers is relatively complex, there has been a limit to achieve both a good power efficiency and a good PSD characteristic over a wide frequency bandwidth.
It is, therefore, a primary object of the present invention to provide an innovative composite transmitter having a good power efficiency as well as a preferred PSD over a wide range of frequency, by use of a power combiner which has a symmetric structure for combining the outputs of a first and a second power amplifiers, a fixed-impedance structure, and capability of differentiating the frequencies of the EIS and the main signal in setting up an independent impedance for each of a main signal and an EIS (in contrast to a conventional composite transmitter having a variable impedance for a main signal and an EIS).
Referring to
(1) A device 80 which, upon receipt of an in-phase component of a baseband modulation signal I(t) (the in-phase component signal hereinafter referred to as in-phase signal I) and an orthogonal component Q(t) of the baseband signal, orthogonal to the in-phase signal I (the orthogonal component hereinafter referred to as orthogonal signal Q), generates
a first composite signal S1 in the form of a vectorial sum of
a second composite signal S2 in the form of a vectorial difference between the main signal and the EIS (the device will be hereinafter referred to simply as signal component separating device 80).
(2) A first power amplifier 50 for power amplifying the first composite signal S1 and a second power amplifier 51 for power amplifying the second composite signal S2, and
(3) A power combiner 40 for combining an output of the first power amplifier 50 and an output of the second power amplifier 51 to provide a composite power at its power combining end.
The composite transmitter 201 has a first feature that the signal component separating device is configured to differentiate instantaneous frequencies of the EIS and the main signal by restraining the phase of the EIS within a predetermined range centered about an arbitrarily given phase θ0. The composite transmitter 201 also has a second feature that a first line length between a first input end and the output end of the power combiner 40 and a second line length between a second input end and the output end of the power combiner 40 are such that each line is equivalent to an open circuit for the EIS when the line is viewed from the respective output ends of the first and second power amplifiers.
In a second aspect of the invention, the compound transmitter 201 is configured in the form of a composite transmitter 202 (not shown) in which the signal component separating device provides:
a phase determination signal u(t) which is +1 if the phase of the main signal is in a range between θ0−π/2 and θ0+π/2 but is −1 otherwise;
an in-phase baseband signal Iz of the EIS in the form of a product of
an orthogonal baseband signal Qz of the EIS in the form of a product of
In a third aspect of the invention, the composite transmitter 201 is modified in the form of a composite transmitter 203 (not shown) in which the signal component separating device provides:
a phase determination signal u(t) which equals +1 if the main signal has a phase in the range between θ0−π/2 and θ0+π/2, but −1 otherwise;
an EIS in the form of a product of
the phase determination signal u and
a composite signal composed of
In a fourth aspect of the invention, the composite transmitter 201 is configured in the form of a composite transmitter 204 (not shown) in which,
wherein the signal component separating device provides:
an EIS in the form of a product of
a signal obtained from an envelope signal E, which is obtained from the main signal by replacing the envelope level of the main signal with the peak envelope level C, by subtracting the main signal from the envelope signal E.
In a fifth aspect of the invention, the composite transmitter 201 is given in the form of a composite transmitter 205 shown
where
In a sixth aspect of the invention, the composite transmitter 201 is configured in the form of a composite transmitter 206 (not shown), wherein the signal component separating device provides the EIS such that the following conditions are satisfied:
In a seventh aspect of the invention, an inventive composite transmitter 207 shown
controls gain(s) of a first line between the input end of the signal component separating device and the output end of the transmitter via the first power amplifier and/or a second line between the input end of the input end of the signal component separating device and the output end of the transmitter via the second power amplifier, by means of the gain control signal, whereby reducing ACLR contained in the output signal of the transmitter to a predetermined level.
It is noted that the composite transmitter 201 differentiates the frequencies of a main signal and an efficiency improving signal (EIS) and that the power combiner 40 is structured in a symmetrical configuration to combine the powers of the two power amplifiers 50 and 51, so that output impedances of the main signal and the efficiency improving signal can be set up independently. As a result, the composite transmitter 201 enables operations of the two power amplifiers 50 and 51 with nearly the maximum power efficiency over a wide range of frequency and a wide range of input signal level, in contrast to the Chireix transmitters 101 and the Doherty transmitters 103 which can maximize their power efficiencies only at center frequencies and limited levels for a given input signal.
Referring to
(1) A signal component separating device 80 which, upon receipt of an in-phase baseband signal I(t) and an orthogonal baseband signal Q(t), outputs
(2) A first power amplifier 50 for power amplifying the first composite signal S1(t) and a second power amplifier 51 for power amplifying the second composite signal S2(t); and
(3) A power combiner 40 having a symmetric structure for combining outputs of the power amplifiers 50 and 51 upon receipt of the outputs via respective impedance inverters 60 and 61, providing a composite output signal of the transmitter. It is noted that each of the impedance inverters 60 and 61 is a line having a length which is equivalent to an open circuit for the EIS when the lines are viewed from either output end of the power amplifier 50 or 51.
In conventional composite transmitters (e.g. Chireix transmitter 101 and Doherty transmitter 103), a main signal and an EIS are synchronized (with their phases being either 90 degrees or −90 degrees in the Chireix transmitter 101 and either 0 or 180 degrees in the Doherty transmitters 103 and 104). In contrast, in the composite transmitter 201 of the present invention, output impedances of the power amplifiers 50 and 51 can be set up independently for the main signal and the EIS by differentiating the frequencies of the main signal and the EIS.
In constructing the signal component separating device 80, there are two approaches available, one shown in
In the second approach shown in
Let us call the composite transmitter 201 “resistor-terminated composite transmitter” when the power combiner 40 is replaced by a 180-degree hybrid circuit in such a way that the main signal is output from a zero-degree output port while high-frequency main signal current and the EIS current are passed through respective resistive loads. The DC power consumption P of the inventive composite transmitter 201 is given by the following equation.
P=p0−P0·P2/(P1+P2)=P0·P1/(P1+P2) (Eq. 13)
where
P0 is the d.c. power consumption by a resistor-terminated composite transmitter,
P1 is the high-frequency power of the main signal, consumed by a resistive load connected to the 0-degree port, and
P2 is the high-frequency power of the EIS, consumed by a resistive load connected to the 180-degree output port. Note in this case that the d.c. power consumption P is equal to P0 minus P0 times the ratio of P2/(P1+P2).
Defining by <f(t)> a long-time average of f(t) or magnitude of a direct current, and denoting by η1 and η2 instantaneous power efficiencies of the power amplifiers 50 and 51, respectively, the average power efficiency η of this composite transmitter 201 is given by the following equation in terms of the ratio between the average high-frequency output power and the average dc power.
η=2∫|a|2dt/∫(|a+z|2/η1+|a−z|2/η2)|a|2/(|a|2+|z|2)dt (Eq. 14)
where the integration is made over a time interval such that η may vary within a predetermined range. When the power amplifiers 50 and 51 are both B-class amplifiers having a maximum power efficiency η0, efficiencies η1 and η2 are given by η0 multiplied by the respective normalized envelope levels.
η1=η0|a+z|/C (Eq. 15)
Plugging Eqs. 15 and 16 in Eq. 14, the average power efficiency η of the composite transmitter 201 is obtained from the following equation for the case where the power amplifiers 50 and 51 are B-class amplifiers having a maximum power efficiency η0.
η=2η0∫|a/C|2dt/∫(|a+z|/C+|a−z|/C)|a|2/(|a|2+|z|2)dt (Eq. 17)
The present invention can be embodied in different modes such as embodiments 2 through 6, depending on how EIS is generated. The invention will now be described with reference to these embodiments, along with their power efficiency characteristics and PSD characteristics. If the composite transmitter 201 has a difference in gain between the two power amplifiers 50 and 51, then the adjacent channel leakage ratio (ACLPR) characteristic of the transmitter will become inferior as is the conventional Chireix transmitter 101. However. this problem can be circumvented by configuring the composite transmitter in a manner as embodied in a seventh embodiment.
The composite transmitter 202 (not shown) is a modified version of a conventional Chireix transmitter (
a phase determination signal u(t) (simply referred to as u) which equals +1 if the main signal has a phase in the range between θ0−π/2 and θ0+π/2 but equals −1 otherwise, where θ0 is an arbitrarily set phase, and provides an EIS (Eq. 4), whose vectorial form is given by the following equation
jb=jb(t)exp{φ(t)} (Eq. 18)
where b(t)=√{square root over (C2−a(t)2)}, and C is the peak envelope level of the main signal.
A first and a second vectorial composite signals S1 and S2, respectively, given by Eqs. 19 and 20, respectively.
S1=a+jub (Eq. 19)
S2=a−jub(Eq. 20)
Denoting the main signal by I+jQ and the EIS by Iz+jQz, and defining the quantity b(t)/a(t), or √{square root over (C2/a(t)2−1)} as envelope conversion signal, the in-phase signal Iz of the EIS turns out to be a product of the envelope conversion signal, −Q, and the phase determination signal u, and the orthogonal signal Qz of the EIS turns out to be a product of the envelope conversion signal, the in-phase signal I, and the phase determination signal u.
When the instantaneous frequency of the main signal does not match the carrier frequency fc, the main signal appears to rotate on the phase plain referenced to the carrier frequency fc. In the conventional Chireix transmitters 101 and 102, the EIS also rotates in synchronism with the main signal and has a broader bandwidth than the main signal. The inventive composite transmitter 202 has a feature that the frequencies of the main signal and the EIS are differentiated by restraining the EIS in the form of jub or −jub within one half of the phase plane constructed with reference to a vector having the carrier frequency fc.
u(t)=sign{Q(t)}=Q(t)/|Q(t)| (Eq. 21)
Alternatively, in order to constrain the EIS, −jub, in the left half plane, upper half plane, and lower half plane, respectively, it suffices to choose θ0 to be −π/2, π, or 0, respectively.
In PSD simulations that follow, the main signal is not an unmodulated signal but is a typical signal having a large peak-average power ratio (PAPR), obtained by peak clipping multi-carrier signals (e.g. 4-wave QPSK signal) distributed at equal angular frequency intervals on the frequency axis. The peak clipped signals will be hereinafter referred to as standard RF modulated signal. the main signal is not an unmodulated signal but is an standard RF modulated signal which is a typical multi-carrier signal having a large PAPR (peak level to average power ration). More particularly, the standard RF modulated signal is a 4-wave multicarrier signal having four QPSK signals distributed at equal angular frequency intervals on the frequency axis and subjected to peak clipping. It should be understood, however, that the use of such standard RF modulated signal in simulations is not meant to limit the multi-carrier signal to a 4-carrier signal, nor limit the inventive modulation to QPSK modulation. Therefore, the invention can be applied to a transmitter that employs a general RF modulated signal having a relatively large peak-average power ratio (PAPR), including a 1024-wave OEDM signal.
When the power amplifiers 50 and 51 are B-class amplifiers and their maximum power efficiencies are η0 (=π/4), the power efficiency of the composite transmitter 202 is found to be η0 by plugging z=jub in Eq. 17, irrespective of the voltage of the main signal (
The inventive composite transmitter 203 (not shown) is a modification of the Doherty transmitter 104 obtained by replacing the Doherty's power combiner 142 with a structurally symmetric power combiner 40. With the EIS vector a1 given by Eq. 22,
The first composite signal S1 which is a vectorial sum of the main signal, the phase determination signal u, and the EIS and the second composite signal S2 which is a vectorial difference between the main signal and the EIS are given by the following Eqs. 23 and 24, respectively.
S1=a+ua1 (Eq. 23)
S2=a−ua1 (Eq. 24)
In order to narrow the frequency bandwidth of the EIS of the composite transmitter 203, it is also important for an improvement of its ACLR characteristic to reduce the spectral regrowth power density.
Assuming that the power amplifiers 50 and 51 of the composite transmitter 203 are B-class amplifiers having a maximum power efficiency of η0 (=π/4), the power efficiency of the composite transmitter is obtained by plugging z=ua1 in Eq. 17. The power efficiency is shown in
The composite transmitter 204 (not shown) is a modification of the composite transmitter 203, modified to improve the power efficiency of the latter. The composite transmitter 204 utilizes an EIS a1 that is obtained by replacing the envelope level of the main signal with the peak level C of the main signal minus the main signal, irrespective of the input voltage of the main signal, as calculated by the following equation.
a1={C/|a|−1}a (Eq. 25)
This is a contrast to the composite transmitter 203 in which the EIS a1 assumes different values depending on whether the voltage of the main signal is smaller than C/2 or greater than C/2.
Since the envelope level of either the first composite signal S1 or the second composite signal S2 is equal to C or close to C even in the small power domain when the EIS is given by Eq. 25, the composite transmitter 204 has an a better power efficiency than the composite transmitter 203. The power efficiency of the composite transmitter 204 can be obtained by plugging z=ua1 in Eq. 17, which is represented by curve c183 as shown in
In the foregoing examples described above (composite transmitters 202, 203, and 204), the frequencies of the main signal and the EIS are differentiated by limiting the phase of the EIS within a range between θ0−π/2 and θ0+π/2 for an arbitrarily chosen phase θ0. In contrast, a composite transmitter 205 of a fifth embodiment (
It will be recalled that in the composite transmitters 202, the phase of the EIS is either φ(t)±π/2 and in the composite transmitter 203 and 204, the phase of EIS is φ(t) or φ(t)±π, where φ(t) is the phase of the main signal (Eq. 2), and that the EIS vector begins to rotate rapidly on the phase plane as the instantaneous frequency of the main signal departs from the carrier frequency fc so that the main signal vector rotates on the phase plane referenced to the carrier frequency fc, which results in broadening of the frequency bandwidth of the EIS. Considering this point, in the composite transmitter 205, an inverse rotational angle −φ(t) is added to the main signal to fix the EIS on the phase plane.
With such an opposite rotational angle added to the main signal to thereby fix the phase of the EIS on the phase plane, there will be generated a difference in angular frequency between the main signal and the EIS given by a time derivative dφ(t)/dt, so that the main signal and the EIS will become orthogonal to each other, thereby satisfying one of the criteria for establishing the composite transmitter 201. If the lengths of impedance inverters 60 and 61 are chosen such that the power combining end p0 (
Although the phase θ1 of the EIS (or z) is arbitrary, it is assumed here to be zero for sake of simplicity. The EIS z, first composite signal S1, and second composite signal S2 can be graphically represented on the phase plane referenced to the carrier frequency, as shown in
z=z1(t) (Eq. 26)
S1=a+z (Eq. 27)
S2=a−z (Eq. 28)
Solving these equations for z1(t) under the condition that the first composite signal S1 or the second composite signal S2 has an envelope level equal to the peak envelope level C of the main signal A, z1(t) is given by Eq. 29 below.
z1(t)=√{square root over (C2−Q(t)2)}−|I(t)| (Eq. 29)
Taking arbitrarily the phase θ1 of EIS Z, and writing the EIS in the following form,
z=z2(t)exp(jθ1) (Eq. 30)
The envelope level z2(t) of the EIS is given by
z2(t)=√{square root over (C2−Q′(t)2)}−|I′(t)| (Eq. 31)
where I′(t) is the real part of the main signal shifted in the negative direction by a phase angle θ1, and Q′(t) is the imaginary part of the main signal, as given by the following equations.
I′(t)=I(t)cos θ1+Q(t)sin θ1 and
Q′(t)=Q(t)cos θ1−I(t)sin θ1
Except for cases where θ1 equals 0, π, π/2, or −π/2, the baseband signal of the EIS given by Eq. 31 is rather complex, but the power of the transmitter is not improved at all. Therefore, it is sensible from a practical point of view to chose the phase of the EIS as 0 or π/2.
The composite transmitter 206, a sixth embodiment according to the invention, has a feature that during a period of when either one of the first and the second power amplifier is in operation at a discrete level below the saturated envelope level C (for example C/2, C/3, or 2C/3), the other one is in operation at its saturated level C. Since in any of the foregoing composite transmitters 202 through 205 (embodiments 2 through 5) the frequency bandwidth cannot be made infinitely large due to a limitation on the operational speed of the digital signal processing device used, the first and second power amplifiers are required to have a predetermined linearity in order to lower, below a predetermined level, the level of spectral regrowth that takes place outside the frequency band of the EIS, as described above in the connection with
Is−jQs,−Is−jQs,−Is+jQs, and Is+jQs, (Eq. 32)
Referring to
Is=cos φ−x=(1−d2)/4/x
QS=sin φ=√{square root over (1−cos2φ)}=√{square root over (1−{(1−d2)/4/x+x}2)} (Eq. 33)
where φ is the angle between the vectors OA and OE and x is the normalized envelope level of the main signal.
From the condition that Qs be a real number, the allowable range of x is given by Eq. 34.
1−{(1−d2)/4/x+x}2>0 or ½−d/2≦x≦½+d/2 (Eq. 34)
From the expressions of four candidate EISs on the premise that the phase of the main signal is zero (
exp(jφ)=I/√{square root over (I2+Q2)}+jQ/√{square root over (I2+Q2)} (Eq. 35)
We have discussed in Embodiment 5 how spectral broadening of the EIS can be suppressed by choosing an appropriate one of the four candidates for the EIS in order to have the phase of the EIS fixed at a constant level (zero level, for example). We now follow the same principle in selecting a proper EIS among the four candidates. One way to suppress spectral broadening of the EIS is to select one candidate EIS that has a phase closest to zero phase irrespective of the quadrant of the phase plane in which the main signal vector is located.
{sign(I)Is−jsign(Q)Qs}exp(jφ) (Eq. 36)
The composite transmitter 206 will be referred to as standard composite transmitter 206 when the transmitter uses the same EIS as the transmitter 205, but the first and the second composite signals assume one of binary levels C and C/2 and the main signal has a normalized envelope level not more than ¼ or not less than ¾.
When a multiplicity of discrete envelope levels are available to the first and the second composite signals for a given x (normalized envelope level of the main signal) within a certain range, normalized envelope levels of the first and the second composite signals can take any two different envelope levels (e.g. d1, d2 for example). For example, if d1=⅓ and d2=⅔ and x is between ⅓ and ⅔, the normalized envelope levels of the first and the second composite signals can take d1 and/or d2. In such a case as described above, as is the case of the composite transmitter 205, the most natural and practical selection of good EIS is one having the least phase (that is, close to zero degree).
It has been pointed in the foregoing discussions of the composite transmitters 201-206 that deterioration of ACLR characteristic (adjacent channel leakage-power ratio) will arise if there is a difference in voltage gain between the two power amplifiers 50 and 51. A solution for this problem will now be described below in conjunction with a seventh embodiment according to the invention. A composite transmitter 207 shown in
(1) a structure (e.g. a directional coupler 70 in the example shown in
(2) a structure for outputting a distorted signal that is obtained by cancelling out the main signal from a part of the output S0 and orthogonally demodulating the resultant partial output;
(3) a structure for outputting a gain control signal that is obtained by suppressing the fluctuating amplitude of the distorted signal;
(4) a structure for controlling, by means of the gain control signal,
gain of a first line between an input end of the signal component separating device 80-6 and an output end p0 of the transmitter via the power amplifier 50 and/or
gain of a second line between the input end and the output end p0 via the power amplifier 51.
The gain control of the first and/or second line(s) by the gain control signal suppresses the level of a distorted component to a predetermined level. Sources of distorted signals created in a composite transmitter can be classified into two categories: (i) distorted signals due to non-linear input-output characteristics of the first and the second power amplifiers, (ii) residual distorted signals appearing in the output of the transmitter due to a difference in gain between the first and the second power amplifiers.
Of these distorted signals, the residual distorted signals can be suppressed in a relatively simple manner as described below with reference to
| Number | Date | Country | Kind |
|---|---|---|---|
| 2013-147983 | Jun 2013 | JP | national |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/JP2014/067993 | 6/27/2014 | WO | 00 |
| Publishing Document | Publishing Date | Country | Kind |
|---|---|---|---|
| WO2014/208779 | 12/31/2014 | WO | A |
| Number | Name | Date | Kind |
|---|---|---|---|
| 5264807 | Okubo | Nov 1993 | A |
| 5381153 | Saito | Jan 1995 | A |
| 6285251 | Dent | Sep 2001 | B1 |
| 6311046 | Dent | Oct 2001 | B1 |
| 6697436 | Wright | Feb 2004 | B1 |
| 6751265 | Schell | Jun 2004 | B1 |
| 6798843 | Wright | Sep 2004 | B1 |
| 7515652 | Jensen | Apr 2009 | B2 |
| 7657238 | Grange | Feb 2010 | B2 |
| 7733978 | Lin | Jun 2010 | B2 |
| 8031804 | Sorrells | Oct 2011 | B2 |
| 8081935 | Liang | Dec 2011 | B2 |
| 8384476 | Draxler | Feb 2013 | B2 |
| 8660208 | Brillant | Feb 2014 | B2 |
| 8670732 | Borodulin | Mar 2014 | B2 |
| 8824594 | Mahoney | Sep 2014 | B2 |
| 8942652 | Khlat | Jan 2015 | B2 |
| 8964716 | Jones | Feb 2015 | B2 |
| 8964821 | Coan | Feb 2015 | B2 |
| 9240761 | Reyland, Jr. | Jan 2016 | B1 |
| 20030076166 | Hellberg | Apr 2003 | A1 |
| 20030198300 | Matero | Oct 2003 | A1 |
| 20050152469 | Fusco | Jul 2005 | A1 |
| 20060099919 | Sorrells | May 2006 | A1 |
| 20080002764 | Luu | Jan 2008 | A1 |
| 20080068078 | Iwasaki | Mar 2008 | A1 |
| 20080191801 | Kim | Aug 2008 | A1 |
| 20090180530 | Ahn | Jul 2009 | A1 |
| 20100158155 | Borkar | Jun 2010 | A1 |
| 20100225390 | Brown | Sep 2010 | A1 |
| 20110051842 | van der Heijden | Mar 2011 | A1 |
| 20120105147 | Harris | May 2012 | A1 |
| 20120155572 | Kim | Jun 2012 | A1 |
| 20130016795 | Kunihiro | Jan 2013 | A1 |
| 20130148760 | Hezar | Jun 2013 | A1 |
| 20140118063 | Briffa | May 2014 | A1 |
| 20150103952 | Wang | Apr 2015 | A1 |
| 20150188504 | Kesson | Jul 2015 | A1 |
| 20150195118 | Yan | Jul 2015 | A1 |
| 20150263678 | Kunihiro | Sep 2015 | A1 |
| Number | Date | Country |
|---|---|---|
| 2003536312 | Dec 2003 | JP |
| 2005117599 | Apr 2005 | JP |
| 0205421 | Jan 2002 | WO |
| Entry |
|---|
| International Search Report for application PCT/JP2014/067993 dated Sep. 9, 2014. |
| Number | Date | Country | |
|---|---|---|---|
| 20160164553 A1 | Jun 2016 | US |
