Adaptive linearizer for RF power amplifiers

Details Adaptive linearizer for RF power amplifiers Adaptive linearizer for RF power amplifiers

1999-05-05

2001-03-27

Mottola, Steven J. (Department: 2817)

Amplifiers

Hum or noise or distortion bucking introduced into signal...

C330S151000

Reexamination Certificate

active

06208207

ABSTRACT:

BACKGROUND OF THE INVENTION
This application pertains to improvements in linearization of radio frequency (RF) amplifiers to reduce the effects of intermodulation (IM) distortion.
All amplifiers are non-linear to some degree. If the signal carried by the amplifier has an envelope that fluctuates in magnitude, such as a multicarrier signal or a linear data modulation, then the non-linear operation generates intermodulation (IM) products in the amplifier output. These IM products represent unwanted interference in the operating band of the amplifier. Although it is possible to reduce the power of the IM products relative to the power of the desired signal by reducing the drive level of the amplifier, this expedient also reduces the power efficiency of the amplifier. Increasing the linearity of the amplifier by means of external circuitry can be a more efficient alternative.
A number of prior art approaches to this problem are described in U.S. Pat. No. 5,489,875, which is incorporated herein by reference. Some of the best prior art approaches described therein use a feedforward linearizer.
Traditional feedforward linearizers include a signal cancellation circuit and a distortion cancellation-circuit. The signal cancellation circuit has two branches, one of which contains the power amplifier whose output is to be linearized. In particular, the amplifier's output consists of an amplified version of an input signal, plus IM distortion. The other branch of the signal cancellation circuit contains circuitry characterized by a coefficient &agr; (amplitude and phase) that can be adjusted to match the amplitude and phase shift of the amplifier, and a delay, also chosen to match the amplifier. If the match is perfect, the error signal obtained by subtracting the output of the two branches of the signal cancellation circuit equals the IM distortion. In the distortion cancellation circuit, an appropriately amplified and phase shifted version (coefficient &bgr;) of the distortion is subtracted from the amplifier output, ideally leaving only the linearly amplified replica at the feedforward output.
FIG. 1
shows an example of a traditional prior art feedforward amplifier. The incoming signal is split by splitter S
1
into two paths comprising the signal cancellation circuit. The first path
10
,
15
,
20
contains a complex gain adjuster CGA
1
and the main amplifier A
1
, the output
20
of which contains the amplified desired signal and unwanted IM distortion. Splitter S
2
directs part of the main amplifier output along line
25
to combiner C
1
. The second path
30
,
35
,
40
carries the desired signal, delayed by delay line DL
1
to match the delay in the first path, to another input of combiner C
1
. The complex gain adjuster CGA
1
provides means to change the amplitude and phase so that the signal component is cancelled in combiner C
1
, leaving only the IM distortion at line
45
. The distortion cancellation circuit also consists of two branches. In one, the IM distortion on line
45
passes through complex gain adjuster CGA
2
and auxiliary amplifier A
2
to combiner C
2
, which receives at its other input
70
the main amplifier output, delayed in delay line DL
2
to match the delay of path
25
,
45
,
50
,
55
,
60
. When complex gain adjuster CGA
2
is correctly adjusted, the IM distortion is cancelled in combiner C
2
, leaving only the amplified input signal at its output
75
.
Typical implementations of the complex gain adjuster are shown for polar coordinates in FIG.
2
(
a
) and for rectangular coordinates in FIG.
2
(
b
). The input, output and two components of complex gain are denoted by I, O, GA and GB, respectively.
The complex gain adjuster CGA
1
can alternatively be placed in line
30
, although doing so precludes cancellation of any distortion introduced by the complex gain adjuster itself.
Because feed forward linearization is based on subtraction of nearly equal quantities, its major parameters must adapt to changes in operating environment, such as signal level, supply voltage and temperature.
The “minimum power” principle may be implemented in the prior art feedforward amplifier of FIG.
1
. In the signal cancellation circuit, controller CT
1
operates to minimize the power measured on line
100
using control lines
110
and
115
to complex gain adjuster CGA
1
. This approach does not make use of line
105
. Instead, the system increments the voltages on control lines
110
,
115
in the direction that results in a lower power measured on line
100
.
The “gradient” method is an alternative to the minimum power principle for adaptation.
FIG. 3
shows that the signal cancellation controller CT
1
is a bandpass correlator. The signal for which the power is to be minimized at input I and a reference signal at input R are split in splitters S
101
, S
102
, respectively, and one of them is phase shifted by 90 degrees in phase shifter PS
1
. Two bandpass mixers M
101
, M
102
produce outputs for which the mean value indicates the direction and size of increments to the complex gain components. Integrators I
1
, I
2
remove high frequency noise and sum the increments to produce the complex gain components at outputs GA and GB. The controller therefore operates to bring the mean value of the gradient to zero. Numeric designations on the input and output lines indicate where the bandpass controller is connected in the signal cancellation circuit. Other embodiments of the gradient method adapt the control voltages to complex gain adjuster CGA
1
similarly. The gradient method is faster than previously proposed minimum power methods and does not require deliberate misadjustments in order to determine the direction of change. However, it is sensitive to DC offset at the output of the mixers that create the gradient signal.
The gradient method can also be applied to adaptation of the
FIG. 1
distortion cancellation circuit, as indicated in
FIG. 3
by the numeric designations in parentheses. Specifically, controller CT
2
operates to bring the mean value of the correlation between the signal on line
85
and the signal on line
95
to zero using control lines
120
and
125
to complex gain adjuster CGA
2
.
A number of more sophisticated approaches are also disclosed in the '875 patent. In one of these approaches, the delay, gain and phase differences are automatically adjusted according to a gradient principle, instead of merely adjusting the gain and the phase.
The '875 patent also discloses approximating the gradient as a sum of partial gradients taken over limited bandwidths. In the case of the distortion cancellation circuit, this allows calculation of the gradient over selected frequency bands that do not contain the amplified input signal, in order to reduce the masking effect. The use of limited bandwidth for each partial gradient allows use of digital signal processing technology to perform the calculation, thereby eliminating the DC offset that could otherwise cause convergence to an incorrect value.
The '875 patent also discloses automatically adjusting the differences to minimize the power at the output of the corresponding cancellation circuit. At each adjustment step, a set of measurements corresponding to perturbed values of the parameters (describing delay, gain and phase) is made. From these measurements, an estimate of the gradient of the power surface is formed. All the parameters describing delay, gain and phase are then adjusted in a direction opposite to the gradient, thereby effecting the greatest decrease in the power to be minimized. In the case of the distortion cancellation circuit, the power to be minimized is the sum of powers measured in selected frequency bands that do not contain the amplified input signal, in order to reduce the masking effect.
FIG. 4
depicts another prior art feed forward amplifier that is disclosed in the '875 patent. The input signal on line
5
enters the signal cancellation circuit, where splitter S
1
produces two branches. The upper branch consists of the delay, gain and phase adjust

Affiliated with

Also associated with

Rating

Adaptive linearizer for RF power amplifiers does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Adaptive linearizer for RF power amplifiers, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Adaptive linearizer for RF power amplifiers will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2440476