Amplifiers – Hum or noise or distortion bucking introduced into signal...
Reexamination Certificate
2002-06-07
2004-01-20
Shingleton, Michael B (Department: 2817)
Amplifiers
Hum or noise or distortion bucking introduced into signal...
C330S151000
Reexamination Certificate
active
06680649
ABSTRACT:
FIELD OF THE INVENTION
The present invention pertains to linear amplification of RF signals, for example linear amplification of RF signals using a multicarrier amplifier.
RELATED ART AND OTHER CONSIDERATIONS
Amplifiers are typically employed to amplify RF signals in order to provide, e.g., increased power for transmission purposes, particularly transmission over an air interface to a receiver such as (for example) a mobile station (e.g., a user equipment unit (UE) such as a cell phone). But in amplifying an input RF signal, the amplifier may add unwanted components due to non-linear characteristics of the amplifier. Such is particularly true when the type of amplifier utilized is chosen for its power efficiency and/or when plural continuous wave RF input signals are applied to the amplifier. Rather than just producing amplified signals corresponding input signals, such amplifier may also output certain additional signals related to the frequencies of the input signals. In this regard, mathematically the output of the amplifier can be expressed as a DC term; a fundamental term (which includes nominal gain for the input signals and an amplitude distortion); and (typically second and third) harmonics terms. The DC term and harmonics can usually be filtered out rather easily, leaving a passband.
The distortion within the passband is not easily removed, but rather is minimized by designing the overall amplifier system in order to compensate for the non-linear characteristics of the amplifier component per se. Such “linearization” of an amplifier system is important in order to avoid distorted signal trajectories and to avoid errors in determining the logic level of individual digital signals.
There are many techniques that can be used to linearize amplifiers. Among the linearization techniques are the following: Back off (in the case of Class A amplifiers); Feedforward; Vector summation; Predistortion, and Feedback. Several of these linearization techniques are briefly described in U.S. Pat. No. 6,075,411 to Briffa et al., which is incorporated herein by reference in its entirety. See also, in this regard, Briffa, Mark, “Linearisation of RF Power Amplifiers,”, 1996.
The feedforward technique is advantageous for broadband linear RF amplifier systems. As mentioned briefly above, since the multicarrier input signal is distorted by the non-linearities in the main amplifier, certain intermodulation (IM) products appear at the output. In essence, the feedforward technique generates an error signal by comparing the input signal with the main amplifier output. The error signal is subtracted from the main amplifier output, leaving a (nearly) distortion-free amplified signal.
FIG. 1
illustrates a simplified, example amplifier system
20
which employs a feedforward technique to minimize distortion. The amplifier system
20
comprises a phase and gain adjuster
22
which receives, via coupler
24
, an input signal. Output from the phase and gain adjuster
22
is applied to main power amplifier
26
. Output from main power amplifier
26
is applied to a coupler
28
, and from one leg of coupler
28
via attenuator
30
to subtractor
32
. Both subtractor
32
and first loop controller
34
receive, via delay
36
, the input signal as obtained from coupler
24
. Output from subtractor
32
is applied both to first loop controller
34
and to a second gain and phase adjuster
40
. Output from gain and phase adjuster
40
is applied to auxiliary amplifier
42
, whose amplified output is coupled by coupler
44
to line
46
. Line
46
emanates from coupler
28
and delay
48
. The output signal carried on line
46
at point
51
is applied via coupler
50
and attenuator
52
to third loop controller
54
, with third loop controller
54
connected to control gain and phase adjuster
40
.
Being in a simplified form for sake of illustration, the amplifier system
20
of
FIG. 1
comprises three loops. A first loop of amplifier system
20
includes phase and gain adjuster
22
, main power amplifier
26
, coupler
28
, attenuator
30
, and subtractor
32
. If the gain and phase shift through phase and gain adjuster
22
, main power amplifier
26
, and attenuator
30
equals the gain and phase shift through delay
36
, an error signal indicative of the distortion of main power amplifier
26
is output by subtractor
32
. But in order to equalize gain and phase shift through these paths, first loop controller
34
is used to produce control signals, applied on line
60
, to phase and gain adjuster
22
.
A second loop of amplifier system
20
comprises attenuator
30
, subtractor
32
, gain and phase adjuster
40
, auxiliary amplifier
42
, coupler
44
, and delay
48
. If the gain and phase shift through attenuator
30
, subtractor
32
, gain and phase adjuster
40
, and auxiliary amplifier
42
equals the gain and phase shift through delay
48
, except for a 180 degree phase shift, the distortion is added in opposite phase at coupler
44
, thus canceling out the distortion of main power amplifier
26
on line
46
. A third loop including attenuator
52
and third loop controller ensures phase and gain equality in these two paths.
Thus, the first loop described above with reference to amplifier system
20
creates an error signal which contains the intermodulation distortion from the main power amplifier
26
. The second loop serves to cancel intermodulation distortion at output point
51
, while leaving the carriers unaffected.
If the intermodulation distortion could be reduced in the main amplifier, the demands on the feedforward system would be lowered. For example, the error amplifier, e.g., auxiliary amplifier
42
in
FIG. 1
, must be dimensioned to handle the total intermodulation power. With less intermodulation power, a smaller error amplifier with a lower power consumption could be used, leading to higher efficiency of the overall amplifier system.
Accordingly, some amplifier systems which employ the feedforward technique also introduce a predistortion that is the reverse of the non-linearities of the main amplifier.
FIG. 2
illustrates an example use of pre-distortion in a system such as that described above. The system of
FIG. 2
basically involves addition of a pre-distortion circuit
80
, as well as an additional phase and gain adjuster
22
′. The additional phase and gain adjuster
22
′ is added in series with the first gain and phase adjuster
22
(which can be a Cartesian gain and phase adjuster). The person skilled in the art knows how to construct and use a suitable pre-distortion circuit, for example with reference to U.S. Pat. No. 6,075,411 to Briffa et al., which is incorporated herein by reference in its entirety.
At high output levels, the gain of a power amplifier (such as power amplifier
26
in
FIG. 2
) may decrease due to compression. A purpose of predistortion is to add (complex) gain in signal path before the power amplifier in order to compensate for the decrease of gain in the power amplifier.
The pre-distortion gain and phase adjuster
22
′ has the normalized transfer function 1+i
p
(x)+jq
p
(x), where i
p
(x)+jq
p
(x)=P(x) is the pre-distortion vector provided by pre-distortion circuit
80
. The functions i
p
(x) and q
p
(x) are signal dependent and are in most cases polynomials. The combined output (seen at the output of pre-distortion gain and phase adjuster
22
′) is provided by Expression 1.
y
(
t
)=
x
(
t
) ((
i
c
+jq
c
)(1
+i
p
(
x
)+
jq
p
(
x
))) (Expression 1)
It would, however, be convenient if the pre-distortion signal output by pre-distortion circuit
80
were added in the first Cartesian phase and gain adjuster
22
, rather than in pre-distortion gain and phase adjuster
22
′.
FIG. 3
shows such an attempted implementation, with part of the input signal (obtained via coupler
82
) and certain coefficients being used by pre-distortion circuit
80
to generate a pre-distortion vector P. In such case it would appear that the pre-distortion vector P is to be added by an a
Nixon & Vanderhye P.C.
Shingleton Michael B
Telefonaktiebolaget LM Ericsson (publ)
LandOfFree
Coordinate rotation of pre-distortion vector in feedforward... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Coordinate rotation of pre-distortion vector in feedforward..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Coordinate rotation of pre-distortion vector in feedforward... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3242195