808636 1
RADIO-FREQUENCY TRANSMITTER, SUCH AS FOR BROADCASTING
AND CELLULAR BASE STATIONS
CROSS-REFERENCE TO RELATED APPLICATIONS
The subject matter of this application is related to that of U.S. Patent
Application No. 13/228063, by Igor Acimovic, attorney docket reference 810317-USNP,
filed on the same date as the present application, and entitled "RADIOFREQUENCY
CIRCUIT HAVING A TRANSCOUPLING ELEMENT," which is
incorporated herein by reference in its entirety.
BACKGROUND
Field of the Invention
The present invention relates to equipment for telecommunication systems
and, more specifically but not exclusively, to radio-frequency (RF) transmitters and
power amplifiers, and passive RF circuits suitable for use therein.
Description of the Related Art
This section introduces aspects that may help facilitate a better understanding
of the invention(s). Accordingly, the statements of this section are to be read in this
light and are not to be understood as admissions about what is in the prior art or what
is not in the prior art.
A recent trend in the telecommunications industry includes the introduction of
wideband digital-modulation systems, such as the third generation (3G) cellularsystem
wideband code-division-multiple-access (WCDMA) and the fourth generation
(4G) cellular-system orthogonal frequency-division multiple-access (OFDMA). This
trend has had a profound effect on power-amplifier specifications because an RF
power amplifier used in a wideband digital-modulation system needs to properly and
efficiently handle a signal that has a fast-changing envelope, a high peak-to-average
power ratio (PAPR), and a bandwidth that can be tens of megahertz. In addition, for
cost reasons, a single power amplifier is usually configured to amplify multiple
modulated carriers.
808636 2
A typical RF power amplifier is inherently nonlinear, with its gain being a
function of the output-power level. The gain usually decreases when the output
power approaches the saturation level of the amplifier, and the phase of the gain can
either increase or decrease, depending on the type of the active medium. Amplitude
and/or phase distortions in the power amplifier tend to cause the generation of
spurious spectral components often referred to as inter-modulation-distortion (IMD)
products. IMD products are detrimental, for example, because they increase the level
of interference between adjacent RF channels.
SUMMARY
Disclosed herein are various embodiments of a transmitter that uses a digital
pre-distortion (DPD) circuit to mitigate the effects of nonlinearity of a multistage or
multi-branch power amplifier, such as a Doherty power amplifier or a Chireix power
amplifier. The DPD circuit relies on two or more feedback signals received from an
RF-output circuit of the transmitter to generate individually pre-distorted signals for
each stage/branch of the power amplifier. The use of these individually pre-distorted
signals advantageously enables the transmitter to achieve a more efficient suppression
of inter-modulation-distortion products than that typically achieved with a comparable
prior-art transmitter.
According to one embodiment, provided is an apparatus having: a digital predistortion
circuit configured to pre-distort a digital input signal to generate a first predistorted
digital signal and a second pre-distorted digital signal different from the first
pre-distorted signal; a first amplifier branch configured to generate a first amplified
signal based on the first pre-distorted digital signal; a second amplifier branch
configured to generate a second amplified signal based on the second pre-distorted
digital signal; and a radio-frequency (RF) circuit configured to combine the first and
second amplified signals to generate a combined signal. The RF circuit is further
configured to generate first and second feedback signals based on at least two of the
first amplified signal, the second amplified signal, and the combined signal. The
digital pre-distortion circuit is configured to generate the first pre-distorted digital
signal and the second pre-distorted digital signal based on said first and second
feedback signals to counteract nonlinearity in the first and second amplifier branches.
808636 3
According to another embodiment, provided is a signal-amplification method
having the steps of: pre-distorting a digital input signal to generate a first predistorted
digital signal and a second pre-distorted digital signal different from the first
pre-distorted signal; generating a first amplified signal based on the first pre-distorted
digital signal in a first amplifier branch; generating a second amplified signal based
on the second pre-distorted digital signal in a second amplifier branch; combining the
first and second amplified signals in a radio-frequency (RF) circuit to generate a
combined signal; and generating first and second feedback signals based on at least
two of the first amplified signal, the second amplified signal, and the combined signal.
The step of pre-distorting comprises the sub-step of generating the first pre-distorted
digital signal and the second pre-distorted digital signal based on said first and second
feedback signals to counteract nonlinearity in the first and second amplifier branches.
According to yet another embodiment, provided is an apparatus having:
means for pre-distorting a digital input signal to generate a first pre-distorted digital
signal and a second pre-distorted digital signal different from the first pre-distorted
signal; a first amplifier branch configured to generate a first amplified signal based on
the first pre-distorted digital signal; a second amplifier branch configured to generate
a second amplified signal based on the second pre-distorted digital signal; means for
combining the first and second amplified signals to generate a combined signal; and
means for generating first and second feedback signals based on at least two of the
first amplified signal, the second amplified signal, and the combined signal. The
means for pre-distorting is configured to generate the first pre-distorted digital signal
and the second pre-distorted digital signal based on said first and second feedback
signals to counteract nonlinearity in the first and second amplifier branches.
BRIEF DESCRIPTION OF THE DRAWINGS
Other aspects, features, and benefits of various embodiments of the invention
will become more fully apparent, by way of example, from the following detailed
description and the accompanying drawings, in which:
FIG. 1 shows a block diagram of a radio-frequency (RF) transmitter according
to one embodiment of the invention;
808636 4
FIG. 2 shows a block diagram of a transceiver that can be used in the RF
transmitter of FIG. 1 according to one embodiment of the invention;
FIG. 3 shows a circuit diagram of an RF circuit that can be used in the RF
transmitter of FIG. 1 according to one embodiment of the invention;
FIG. 4 shows a circuit diagram of an RF circuit that can be used in the RF
transmitter of FIG. 1 according to another embodiment of the invention; and
FIG. 5 shows a top view of a microstrip circuit that can be used to implement
the RF circuit of FIG. 4 according to one embodiment of the invention; and
FIG. 6 shows a circuit diagram of a two-stage amplifier according to one
embodiment of the invention.
DETAILED DESCRIPTION
One method that can be used to linearize the nonlinear response of a radiofrequency
(RF) power amplifier over its intended dynamic range is digital predistortion
(DPD). DPD works in the digital domain and uses digital-signal-processing
techniques to pre-distort a baseband signal before modulation, up-conversion, and
amplification. With DPD, the power amplifier can be utilized substantially up to its
saturation point while maintaining a sufficiently accurate linear relationship between
the input and output signals. DPD is an attractive technique, e.g., because it can
significantly increase the power efficiency of a power amplifier and be implemented
using standard and/or inexpensive circuit components. A high degree of flexibility
can be achieved if programmable hardware is used, such as digital signal processors
(DSPs) and/or field-programmable gate arrays (FPGAs). In addition, DPD does not
require significant changes in the schematics of the costly analog part (e.g., the RFoutput
circuit) of the corresponding transmitter and lends itself to various
advantageous implementations in which the analog front end of the transmitter has a
relatively small size, and the configurable digital part of the transmitter can be placed
very close to the antenna.
FIG. 1 shows a block diagram of an RF transmitter 100 according to one
embodiment of the invention. Transmitter 100 uses a Doherty amplification scheme
to convert a digital input signal 102 into an analog RF-output signal 152. Output
signal 152 is applied to an output load (e.g., an antenna) 160. A feedback path
808636 5
comprising a feedback-receiver (FBR) circuit 120 and a DPD circuit 110 enables
transmitter 100 to apply digital pre-distortion to input signal 102, thereby suppressing
IMD products in output signal 152.
The Doherty amplification scheme of transmitter 100 employs power
amplifiers 140i and 140 connected in parallel as indicated in FIG. 1. Output signals
142i and 142 generated by amplifiers 140i and 140 , respectively, are combined in
an RF-output circuit 150 as further described below in reference to FIGs. 3-4 to
produce output signal 152. Amplifier 140i is configured to operate, e.g., as a class-B
or a class-AB amplifier and is also referred to as a primary or carrier stage. Amplifier
140 is configured to operate, e.g., as a class-C amplifier and is also referred to as an
auxiliary or peak stage. A brief description of operating configurations corresponding
to the pertinent amplifier classes can be found, e.g., in U.S. Patent Nos. 7,498,876 and
7,928,799, both of which are incorporated herein by reference in their entirety.
Due to different configurations of amplifiers 140i and 140 , only amplifier
140i provides signal amplification when input signal 102 and therefore RF signals
132i and 132 are small. Amplifier 140 remains turned off until RF signal 132
reaches a certain threshold level. Near this threshold level, amplifier 140i is close to
saturation, and amplifier 140 turns on to supply the output-signal portion that tends to
be clipped off by the near-saturation operating regime of amplifier 140i. This
complementary action of amplifiers 140i and 140 enables transmitter 100 to
advantageously have relatively high power efficiency for a wide range of input-signal
levels. A more detailed explanation of how high power efficiency can be achieved for
amplifiers 140i and 140 is provided below in reference to FIG. 3.
In addition to producing output signal 152 by properly combining signals 142i
and 142 , RF-output circuit 150 is configured to generate feedback signals 148i-1483
based on signals 1421 142 , and 152 and supply these feedback signals to FBR circuit
120. In one embodiment, feedback signal 148i provided by RF-output circuit 150 to
FBR circuit 120 is an attenuated copy of signal 142^ feedback signal 148 is an
attenuated copy of signal 142 ; and feedback signal 1483 is an attenuated copy of
signal 152. In various alternative embodiments, RF-output circuit 150 can be
configured to provide to FBR circuit 120 only two of feedback signals 148i-1483
808636 6
and/or generate each of feedback signals 148i-148 3 based on a respective different
linear combination of signals 1421 142 , and 152.
In one embodiment, FBR circuit 120 comprises three feedback receivers (not
explicitly shown in FIG. 1), each configured to process a corresponding one of
feedback signals 148i-148 3. The typical processing performed by a feedback receiver
includes down-converting the corresponding feedback signal 148 to baseband and
applying analog-to-digital conversion to the resulting analog baseband signal to
generate a corresponding digital feedback signal 118. Digital feedback signals 118i-
II8 3 generated by FBR circuit 120 correspond to analog feedback signals 148i-148 3,
respectively.
In general, the gain, efficiency, and AM-PM (amplitude-to-phase modulation)
characteristics (e.g., the insertion-phase change as a function of signal amplitude) of a
power amplifier, such as power amplifier 140i or 140 , are all functions of both the
output power and the load impedance. In a typical prior-art DPD scheme, the
individual stimuli (input signals) for the carrier and peak stages of a Doherty power
amplifier are generated by means of a 3-dB power splitter configured to split the RF
signal that is generated based on a single pre-distorted digital signal generated by the
DPD circuit. This means that the stimuli applied to the carrier and peak stages have a
fixed phase relationship with one another. However, as already indicated above, the
carrier and peak stages of a Doherty power amplifier are configured to operate in
different regimes, which causes their output signals to generally have a phase
mismatch between them. Moreover, this phase mismatch varies over time owing to
the variations in the output-power level. Disadvantageously, this prior-art DPD
scheme is not capable of equalizing the phase mismatch and relies mostly on
amplitude pre-distortion for the suppression of IMD products.
This and other pertinent problems in the prior art are addressed in transmitter
100 by configuring DPD circuit 110 to generate individually pre-distorted digital
signals 112i and 112 for amplifiers 140i and 140 , respectively, based on multiple
digital feedback signals, e.g., two or three of signals 118i-118 3. These feedback
signals provide sufficient information to enable the DPD circuit to implement both
amplitude pre-distortion and phase equalization for amplifiers 140i and 140 . In part
due to a relatively small phase mismatch between output signals 142i and 142 ,
808636 7
transmitter 100 is able to better suppress IMD products in its output signal (i.e., signal
152) than a comparable prior-art transmitter.
In one configuration, DPD circuit 110 uses digital feedback signals 118i-1183
to adaptively pre-distort input signal 102 to generate individually pre-distorted digital
signals 112i and 112 . Pre-distorted digital signal 112i is applied to a transmitter
1301 where it is first converted into a corresponding analog signal (not explicitly
shown in FIG. 1). Transmitter 130i then up-converts this analog signal from
baseband to generate RF signal 132i. Pre-distorted digital signal 112 is similarly
processed in a transmitter 130 to generate RF signal 132 . As already indicated
above, RF signals 132i and 132 are the input signals (stimuli) that are applied to
amplifiers 140i and 140 , respectively.
DPD circuit 110 is configured to generate pre-distorted digital signal 112i by
applying a first nonlinear function to input signal 102, wherein the first nonlinear
function creates an expanding nonlinearity that is an approximate inverse of the
compressing nonlinearity of (e.g., compressive amplitude distortion in) amplifier
140i. DPD circuit 110 is further configured to generate pre-distorted digital signal
112 by applying a second nonlinear function to input signal 102, wherein the second
nonlinear function creates a nonlinearity that is an approximate inverse of the
nonlinearity of amplifier 140 . As already indicated above, the first and second
nonlinear functions usually differ from one another due to different operating
configurations of amplifiers 140i and 140 .
In various alternative configurations, DPD circuit 110 can similarly apply
other types of first and/or second nonlinear functions to input signal 102 to generate
pre-distorted digital signals 112i and 112 . In general, the first and second nonlinear
functions are constructed in an inter-related manner to cause the forward signal path
comprising DPD circuit 110, transmitters 130i and 130 , amplifiers 140i and 140 ,
and RF-output circuit 150 to exhibit substantially linear signal-transfer characteristics.
By substantially linear signal-transfer characteristics, it is meant that, with digital predistortion,
the relationship between output signal 152 and input signal 102 can be
approximated well by a constant gain, e.g., a complex or real gain value that does not
depend on the input- (or output-) signal level within the intended dynamic range of
transmitter 100. Description of representative DPD circuits and algorithms that can
808636 8
be used to implement DPD circuit 110 can be found, e.g., in U.S. Patent Nos.
7,957,707, 7,904,033, 7,822,146, 7,782,979, 7,729,446, 7,606,324, 7,583,754, and
7,471,739, all of which are incorporated herein by reference in their entirety.
FIG. 2 shows a block diagram of a transceiver 200 that can be used in
transmitter 100 (FIG. 1) according to one embodiment of the invention. Note that
DPD circuit 110 shown in FIG. 2 is not part of transceiver 200. Digital-to-analog
converters (DACs) 234 and an I-Q modulator 236 can be used to implement
transmitter 130i or transmitter 130 . Analog-to-digital converters (ADCs) 224 and an
I-Q de-modulator 226 can be used to implement a portion of FBR circuit 120.
Transceiver 200 also has a local-oscillator (LO) source 244 that is configured to
supply a local-oscillator (carrier-frequency) signal 246 to I-Q modulator 236 and I-Q
de-modulator 226. In a representative embodiment, transmitter 100 may have more
than one instance of transceiver 200.
In operation, I-Q de-modulator 226 demodulates feedback signal 148 in a
conventional manner by mixing it with LO signal 246. A resulting baseband signal
225 generated by I-Q de-modulator 226 has two components: an in-phase component
225i and a quadrature-phase component 225Q. Signals 225i and 225Qare analog
signals that are converted into digital form by ADCs 224. The resulting digital
signals ¾, and Q are components of the corresponding one of digital signals 118i-
1183 (also see FIG. 1).
DPD circuit 110 uses digital signals ¾, and Q to determine the amount of
distortion in the forward signal path of transmitter 100 (FIG. 1). For example, a
symbol received by DPD circuit 110 via digital signals ¾, and Q b can be combined
with (e.g., summed with and/or subtracted from) one or more corresponding symbols
received by the DPD circuit via the other one or more digital signals 118 (see FIG. 1).
DPD circuit 110 then uses the corresponding original constellation symbol received
via input signal 102 to determine the amount of pre-distortion that needs to be applied
to the original I and Q components to counteract (e.g., cancel or significantly reduce)
the distortion imposed by the forward signal path. The determined amount of predistortion
can be partitioned, in any suitable manner, into a first portion and a second
portion. The first portion, in form of the first nonlinear function, is applied to input
signal 102 to generate pre-distorted digital signal 112i, while the second portion, in
808636 9
form of the second nonlinear function, is similarly applied to input signal 102 to
generate pre-distorted digital signal 112 . Similar to signals 102 and 118, signal 112
is shown as having two components: an in-phase component labeled IPd and a
quadrature-phase component labeled QPd. Note that FIG. 2 shows the generation of
only one of pre-distorted digital signals 112i and 112 . The other one of these signals
can be similarly generated.
Components IPd and QPd of pre-distorted digital signal 112 are converted, in
DACs 234, into analog signals 235i and 235Q, respectively. I-Q modulator 236 then
uses analog signals 235i and 235Q to modulate LO signal 246. The resulting
modulated carrier signal is RF signal 132 (also see FIG. 1). As already indicated
above, signal 132 generated by I-Q modulator 236 can be one of signals 132i and
132 (see FIG. 1). The other one of these signals can be similarly generated.
FIG. 3 shows a circuit diagram of an RF circuit 300 that can be used as RFoutput
circuit 150 (FIG. 1) according to one embodiment of the invention. RF circuit
300 comprises a transcoupler 310, an impedance transformer 320, a directional
coupler 330, and six terminals labeled A through F. The impedances indicated in
FIG. 3 are exemplary and correspond to an implementation in which each of the
external terminals is intended for being connected to a 50-Ohm line, driver, load, or
terminator. One of ordinary skill in the art will understand how to change the various
impedance values shown in FIG. 3 to match RF circuit 300 to an impedance value
different from 50 Ohm.
In a representative configuration, terminals A-F can be connected as follows.
Terminal A is configured to carry feedback signal 148i (see FIGs. 1 and 2). Terminal
B is configured to carry feedback signal 148 (see FIGs. 1 and 2). Terminal C is
configured to receive amplified signal 142i (see FIG. 1). Terminal D is configured to
receive amplified signal 142 (see FIG. 1). Terminal E is configured to carry
feedback signal 1483 (see FIGs. 1 and 2). Terminal F is configured to apply output
signal 152 to load 160.
This representative configuration can be modified to produce several
alternative configurations. For example, three different alternative configurations can
be obtained by changing the connection of terminal A, B, or E from the above808636
10
indicated to a 50-Ohm terminator. In each of these three alternative configurations,
RF circuit 300 will provide only two of feedback signals 148i-1483.
Transcoupler 310 is a four-terminal device having two parallel branches 312
and 314 that are located sufficiently close to each other for an RF signal propagating
through branch 314 to electromagnetically couple into branch 312. Branch 314
comprises a quarter-wave impedance inverter disposed between terminals C and D.
The signal coupling between branches 312 and 314 is relatively weak, e.g., about -30
dB, which ensures a minimal influence of branch 312 on the operation of the quarterwave
impedance inverter in branch 314.
In operation, the quarter-wave impedance inverter of branch 314 can be used
to implement active load modulation for carrier stage 1401 for example, as follows
(also see FIG. 1).
At low input-signal levels, peak stage 140 is in the off state while carrier stage
1401 acts as a controlled current source. Peak stage 140 (ideally) sees infinite
impedance, and the impedance inverter of branch 314 causes carrier stage 140i to see
a higher than 50-Ohm impedance load. The higher impedance load causes carrier
stage 140i to reach near-saturation when its output current reaches only about one half
of its nominal maximum value. When carrier stage 140i is close to saturation, it
advantageously works with nearly maximum power efficiency.
As soon as the input-signal level becomes sufficiently high to turn on peak
stage 140 , the peak stage begins to apply additional current to terminal D. Peak stage
140 now acts as a controlled current source, and carrier stage 140i acts as a
controlled voltage source. The additional current applied by peak stage 140 to
terminal D causes an increase in the output impedance seen by the quarter-wave
impedance inverter of branch 314. Since the input and output impedances of a
quarter-wave impedance inverter are related to one another as duals, the increase in
the output impedance causes a corresponding decrease in the input impedance. Note
that the input impedance of the quarter-wave impedance inverter of branch 314 is the
load that is seen by carrier stage 140i. As the load of carrier stage 140i decreases, the
output current of the carrier stage increases, with the output voltage remaining close
to the saturation level.
808636 11
As the input-signal level increases further, the output impedance of the
quarter-wave impedance inverter of branch 314 keeps increasing, and the effective
load of carrier stage 140i keeps decreasing. In this manner, the impedance inverter of
branch 314 enables peak stage 140 to modulate the load of carrier stage 140i during
high input-signal levels. The load modulation, in turn, keeps carrier stage 140i
operating in a regime that is characterized by advantageously high power efficiency.
Impedance transformer 320 comprises a length of transmission line that is onequarter
of a wavelength long and has an impedance of about 35 Ohm. (Note that, for
a device that has a desired operating-frequency range, the quarter-wave length
typically corresponds to the center frequency of that operating range.) Since
impedance transformer 320 is terminated at 50 Ohm by directional coupler 330, it
presents at terminal D an input impedance of 25 Ohm. The latter impedance matches
the output impedance of two 50-Ohm transmission lines connected in parallel to
terminal D.
Directional coupler 330 comprises branches 332 and 334. Branch 334
operates to present a fixed 50-Ohm termination to impedance transformer 320. The
signal coupling between branches 332 and 334 is relatively weak, e.g., about -30 dB.
Terminal G is connected to a 50-Ohm terminator 340. Terminal E outputs an
attenuated copy of the RF signal presented at terminal H by impedance transformer
304.
Feedback signals SA, ¾ , and collected at terminals A, B, and E,
respectively, of RF circuit 300 can be used to directly measure effective transfer
functions Ti and T2 of stages 140i and 140 , for example, based on Eqs. (l)-(3):
T =^- (1)
T = ~ (2)
ap
T P + T2p 2 = (3)
c
where is a constant that represents the signal-coupling strength between branches
312 and 314; c is a constant that represents the signal-coupling strength between
branches 332 and 334; and p and ¾ represent pre-distorted signals 112i and 112 ,
respectively. The measurement can be done on line or using an appropriate off-line
808636 12
calibration procedure. In principle, any two of Eqs. (l)-(3) are sufficient for the
determination of transfer functions T and T2, provided that the coupling strengths are
known. The use of all three equations enables the determination of the ratio (ale) of
the coupling strengths and, as such, can be used when only one of the two coupling
strengths is known. After the individual transfer functions of stages 140i and 140 are
determined, their reciprocals can be used in a relatively straightforward manner for
amplitude pre-distortion and phase equalization.
FIG. 4 shows a circuit diagram of an RF circuit 400 that can be used as RFoutput
circuit 150 (FIG. 1) according to another embodiment of the invention. In
terms of its intended function, RF circuit 400 is generally analogous to RF circuit 300
(FIG. 3). Therefore, the above-described connections of terminals A-F apply to RF
circuit 400 as well as to RF circuit 300. However, one difference between RF circuits
300 and 400 is that the latter employs a transcoupler 450 that performs the functions
that are similar to the above-described functions of both impedance transformer 320
and directional coupler 330 (see FIG. 3).
The use of transcoupler 450 can provide one or more of the following
benefits/advantages :
(1) A transmitter (e.g., transmitter 100, FIG. 1) employing RF circuit 400 can have
a relatively high power efficiency, e.g., because transcoupler 450 has a lower
insertion loss than a series consisting of impedance transformer 320 and
directional coupler 330;
(2) RF circuit 400 can have a relatively small size because transcoupler 450
occupies a relatively small area on a printed circuit board (PCB); and
(3) The relatively small size of transcoupler 450 and the improved power
efficiency of the corresponding power amplifier can be leveraged to reduce
per-unit fabrication and operating costs.
As evident from the description of transcouplers 310 and 450 in FIGs. 3 and 4,
a transcoupler is a circuit element that has two branches that can be referred to as the
main branch and the auxiliary branch, respectively. The main branch has a length of
about one quarter of the carrier wavelength and is configured to operate as an
impedance inverter that presents a first impedance at the first end of the branch, said
first impedance being proportional to an inverse of a second impedance presented to
808636 13
the branch at the second end of the branch. If the second impedance is a fixed
impedance, then the first branch operates as a quarter-wave impedance transformer.
Branches 314 and 454 are the main branches in transcouplers 310 and 450,
respectively. The auxiliary branch is electromagnetically coupled to the main branch
and configured to operate as a signal coupler that receives an attenuated copy of a
signal from the main branch. Branches 312 and 452 are the auxiliary branches in
transcouplers 310 and 450, respectively.
FIG. 5 shows a top view of a microstrip circuit 500 that can be used to
implement RF circuit 400 (FIG. 4) according to one embodiment of the invention.
Circuit 500 comprises a dielectric substrate 502 and two conducting layers attached to
the opposite (e.g., top and bottom) sides of the dielectric substrate. Only the patterned
top layer is visible in the view provided in FIG. 5. The various microstrip shapes of
this patterned layer define the circuit elements of circuit 500. The bottom layer
(typically referred to as the "ground plane") is not visible in the view provided in FIG.
5. In a representative embodiment, the ground plane is not patterned and comprises a
continuous layer of metal, such as copper.
Circuit 500 has seven terminals labeled A through G. The terminals labeled
by the same letter in FIGs. 4 and 5 are functionally similar. Therefore, the terminals
of circuit 500 can be electrically connected to external circuits, e.g., as already
described above in reference to FIGs. 3 and 4.
Microstrips 512 and 514 are used to implement a transcoupler 510 that is
analogous to transcoupler 310 (see FIGs. 3 and 4). Microstrip 514 is one-quarter of a
wavelength long. Microstrip 512 is electromagnetically coupled to microstrip 514
using two interdigitated combs. One of the combs is electrically connected to
microstrip 512 and is illustratively shown in FIG. 5 as having two fingers 516. The
other comb is electrically connected to microstrip 514 and is illustratively shown in
FIG. 5 as having four fingers 518. Two mitred bends 508 are used to electrically
connect microstrip 512 to terminals A and B.
Microstrips 552 and 554 are used to implement a transcoupler 550 that is
analogous to transcoupler 450 (see FIG. 4). Microstrip 554 is one-quarter of a
wavelength long. Microstrip 552 is electromagnetically coupled to microstrip 554
using two interdigitated combs. One of the combs is attached to a side of microstrip
808636 14
552 and is illustratively shown in FIG. 5 as having two fingers 556. The other comb
is attached to a side of microstrip 554 and is illustratively shown in FIG. 5 as having
four fingers 558. A mitred bend 508 is used to electrically connect microstrip 552 to
terminal E. Microstrips 516 and 562 are used to electrically connect the ends of
microstrip 554 to terminals D and F, respectively.
Circuit 500 also has an optional shunted stub 560 that is connected to
transcoupler 550 as indicated in FIG. 5. Stub 560 is implemented using a microstrip
566 and a shunt 564 that is located at the distal end of that microstrip. Microstrip 566
is one-quarter of a wavelength long. Shunt 564 comprises one or more electrically
conducting vias in dielectric substrate 502 that electrically connect the distal end of
microstrip 566 to the ground plane, thereby short-circuiting stub 560. One function of
stub 560 is to alter the effective impedance of transcoupler 550 compared to the
impedance that the transcoupler would have without the stub. The altered impedance
advantageously has an imaginary part that is substantially nulled and has a very weak
frequency dependence within the pertinent spectral range around the nominal carrier
frequency.
Note that each of microstrips 506, 512, 514, 552, and 562 has a first specified
width, and each of microstrips 554 and 566 has a second specified width that is
greater than the first. In a representative embodiment, the first and second widths are
selected so that (i) each of the RF-transmission lines represented by microstrips 506,
512, 514, 552, and 562 has an impedance of 50 Ohm and (ii) each of the RFtransmission
lines represented by microstrips 554 and 566 has an impedance of 35
Ohm. One skilled in the art will understand how to select other respective widths for
these two sets of microstrips to obtain other impedance values.
FIG. 6 shows a circuit diagram of a two-stage/branch amplifier circuit 600
according to one embodiment of the invention. Circuit 600 can be used, e.g., to
implement a Chireix amplification scheme. A circuit diagram of the corresponding
transmitter can be obtained, e.g., by replacing amplifiers 140i and 140 and RF-output
circuit 150 in transmitter 100 (FIG. 1) by circuit 600. More specifically, signals/lines
6321 632 , 648i, 648 , and 652 in circuit 600 correspond to signals/lines 1321 132 ,
148i, 148 , and 152, respectively, in transmitter 100.
808636 15
The Chireix amplification scheme of circuit 600 employs power amplifiers
640i and 640 connected in parallel as indicated in FIG. 6. Output signals 642i and
642 generated by amplifiers 640i and 640 , respectively, are combined in an RFoutput
circuit 660 comprising output-matching circuits 644i and 644 and
transcouplers 650i and 650 . Amplifiers 640i and 640 are configured to generate
signals 642i and 642 to be each other's complex conjugates and to have a constant
envelope. After RF-output circuit 660 combines phase-modulated signals 642i and
642 , output terminal D of that circuit has the corresponding amplitude-modulated
signal 652.
Similar to RF-output circuits 300 and 400 (FIGs. 3 and 4), RF-output circuit
660 uses load-impedance modulation to achieve relatively high power efficiency for
the amplifier stages. Transcoupler 650i is configured to invert the load impedance,
which is then transformed by output-matching circuit 644i and presented to amplifier
640i. Transcoupler 650 is similarly configured to invert the load impedance, which
is then transformed by output-matching circuit 644 and presented to amplifier 640 .
To properly convert phase-modulated signals 642i and 642 into amplitudemodulated
signal 652, the deviation from the prescribed complex conjugate-phase
relationship between the two amplifier branches in circuit 600 needs to be as small as
possible. Circuit 600 helps to achieve this result by providing feedback signals 648i
and 648 for use with an appropriate DPD circuit that can be similar to DPD circuit
110 of transmitter 100. Transcoupler 650i is configured to generate feedback signal
648i. Transcoupler 650 is similarly configured to generate feedback signal 648 .
Based on these feedback signals, the corresponding DPD circuit can pre-distort
signals 632i and 632 to achieve relatively accurate phase conjugation at terminal D.
In general, transcoupling elements analogous to transcouplers 450, 550, and
650 can be used in any multistage or multi-branch power-amplifier circuit configured
to combine RF-output signals from two or more branches of amplifying elements and
use feedback-based digital pre-distortion to linearize the overall transfer
characteristics of the power amplifier, e.g., to suppress IMD products in its output
signal. The above-described Doherty and Chireix amplification schemes are just two
representative examples of such multistage power-amplifier circuits. From the
description provided herein, one of ordinary skill in the art will be able to use
808636 16
transcoupling elements instead of conventional RF-circuit elements in various other
circuits. Possible benefits/advantages of such use are already indicated above in
reference to FIG. 4.
As used in this specification, the term "radio frequency" (RF) refers a rate of
oscillation in the range of about 3 kHz to 300 GHz. This frequency may be the
frequency of an electromagnetic wave or an alternating current in a circuit. This term
should be construed to be inclusive of the frequencies used in wireless communication
systems.
While this invention has been described with reference to illustrative
embodiments, this description is not intended to be construed in a limiting sense.
Although RF circuit 400 (FIG. 4) has been described as being implemented
using a microstrip technology, it can also be implemented using any other suitable
technology, e.g., a coaxial technology or a planar (such as stripline, slotline, or planarwaveguide)
technology. Other RF circuits disclosed in this specification can similarly
be implemented using these technologies.
As used in the claims, the term "strip" should be construed to cover any
conducting strip, such as a microstrip or a stripline, in the patterned layer of the
corresponding planar circuit or printed circuit board.
In one embodiment, transcoupler 550 (FIG. 5) can be used in conjunction with
one or more circuit elements that differ from directional coupler 510. For example,
instead of being connected to microstrips 506 and 514, microstrip 554 of transcoupler
550 can be connected at the left (as viewed in FIG. 5) end thereof to a microstrip that
has a width different from the width of microstrip 554 and the width of microstrip
562.
Various modifications of the described embodiments, as well as other
embodiments of the invention, which are apparent to persons skilled in the art to
which the invention pertains are deemed to lie within the principle and scope of the
invention as expressed in the following claims.
The present invention may be implemented as circuit-based processes,
including possible implementation on a single integrated circuit.
808636 17
Unless explicitly stated otherwise, each numerical value and range should be
interpreted as being approximate as if the word "about" or "approximately" preceded
the value of the value or range.
The use of figure numbers and/or figure reference labels in the claims is
intended to identify one or more possible embodiments of the claimed subject matter
in order to facilitate the interpretation of the claims. Such use is not to be construed
as necessarily limiting the scope of those claims to the embodiments shown in the
corresponding figures.
Although the elements in the following method claims, if any, are recited in a
particular sequence with corresponding labeling, unless the claim recitations
otherwise imply a particular sequence for implementing some or all of those elements,
those elements are not necessarily intended to be limited to being implemented in that
particular sequence.
Reference herein to "one embodiment" or "an embodiment" means that a
particular feature, structure, or characteristic described in connection with the
embodiment can be included in at least one embodiment of the invention. The
appearances of the phrase "in one embodiment" in various places in the specification
are not necessarily all referring to the same embodiment, nor are separate or
alternative embodiments necessarily mutually exclusive of other embodiments. The
same applies to the term "implementation."
Throughout the detailed description, the drawings, which are not to scale, are
illustrative only and are used in order to explain, rather than limit the invention. The
use of terms such as height, length, width, top, bottom, is strictly to facilitate the
description of the invention and is not intended to limit the invention to a specific
orientation. For example, height does not imply only a vertical rise limitation, but is
used to identify one of the three dimensions of a three dimensional structure as shown
in the figures. Such "height" would be vertical where the microstrips are horizontal
but would be horizontal where the microstrips are vertical, and so on.
Also for purposes of this description, the terms "couple," "coupling,"
"coupled," "connect," "connecting," or "connected" refer to any manner known in the
art or later developed in which energy is allowed to be transferred between two or
more elements, and the interposition of one or more additional elements is
808636 18
contemplated, although not required. Conversely, the terms "directly coupled,"
"directly connected," etc., imply the absence of such additional elements.
The present inventions may be embodied in other specific apparatus and/or
methods. The described embodiments are to be considered in all respects as only
illustrative and not restrictive. In particular, the scope of the invention is indicated by
the appended claims rather than by the description and figures herein. All changes
that come within the meaning and range of equivalency of the claims are to be
embraced within their scope.
808636 19
CLAIMS
What is claimed is:
1. An apparatus, comprising:
a digital pre-distortion circuit configured to pre-distort a digital input signal to
generate a first pre-distorted digital signal and a second pre-distorted digital signal
different from the first pre-distorted signal;
a first amplifier branch configured to generate a first amplified signal based on the
first pre-distorted digital signal;
a second amplifier branch configured to generate a second amplified signal based
on the second pre-distorted digital signal; and
a radio-frequency (RF) circuit configured to combine the first and second
amplified signals to generate a combined signal, wherein:
the RF circuit is further configured to generate first and second feedback
signals based on at least two of the first amplified signal, the second amplified signal,
and the combined signal; and
the digital pre-distortion circuit is configured to generate the first pre-distorted
digital signal and the second pre-distorted digital signal based on said first and second
feedback signals to counteract nonlinearity in the first and second amplifier branches.
2. The apparatus of claim 1, wherein:
the first amplifier branch comprises:
a first transmitter configured to convert the first pre-distorted digital signal
into a first RF signal; and
a first power amplifier configured to amplify the first RF signal to generate the
first amplified signal;
the second amplifier branch comprises:
a second transmitter configured to convert the second pre-distorted digital
signal into a second RF signal;
a second power amplifier configured to amplify the second RF signal to
generate the second amplified signal.
808636 20
3. The apparatus of claim 2, wherein the first power amplifier and the second
power amplifier are configured to operate as a carrier stage and a peak stage,
respectively, of a Doherty power amplifier.
4. The apparatus of claim 2, wherein the first power amplifier and the second
power amplifier are configured to operate as respective phase-modulating stages of a
Chireix power amplifier.
5. The apparatus of claim 1, wherein:
the RF circuit is further configured to generate a third feedback signal; and
each of the first, second, and third feedback signals is an attenuated copy of a
respective one of the first amplified signal, the second amplified signal, and the
combined signal.
6. The apparatus of claim 1, further comprising:
a feedback-receiver circuit configured to:
down-convert the first feedback signal to generate a first digital baseband
signal;
down-convert the second feedback signal to generate a second digital
baseband signal; and
apply said first and second digital baseband signals to the digital pre-distortion
circuit, wherein the digital pre-distortion circuit is configured to generate the first predistorted
digital signal and the second pre-distorted digital signal based on said first
and second digital baseband signals; and
an antenna configured to emit an electromagnetic wave corresponding to the
combined signal.
7. The apparatus of claim 1, wherein the digital pre-distortion circuit is
configured to counteract said nonlinearity by counteracting:
compressive amplitude distortion in at least one of the first and second amplifier
branches; and
a phase mismatch between the first and second amplifier branches.
808636 2 1
8. The apparatus of claim 1, wherein the RF circuit comprises a first directional
coupler, said first directional coupler comprising:
a first branch connected between an output port of the first amplifier branch and
an output port of the second amplifier branch; and
a second branch electromagnetically coupled to the first branch and having first
and second terminals, wherein the first feedback signal appears on the first terminal.
9. The apparatus of claim 8, wherein the RF circuit further comprises a
transcoupler, said transcoupler comprising:
a respective first branch connected in series with the first branch of the first
directional coupler; and
a respective second branch electromagnetically coupled to said respective first
branch and having third and fourth terminals, wherein the second feedback signal
appears on the third terminal.
10. A signal-amplification method, comprising:
pre-distorting a digital input signal to generate a first pre-distorted digital signal
and a second pre-distorted digital signal different from the first pre-distorted signal;
generating a first amplified signal based on the first pre-distorted digital signal in a
first amplifier branch;
generating a second amplified signal based on the second pre-distorted digital
signal in a second amplifier branch;
combining the first and second amplified signals in a radio-frequency (RF) circuit
to generate a combined signal; and
generating first and second feedback signals based on at least two of the first
amplified signal, the second amplified signal, and the combined signal, wherein the
step of pre-distorting comprises generating the first pre-distorted digital signal and the
second pre-distorted digital signal based on said first and second feedback signals to
counteract nonlinearity in the first and second amplifier branches.