U.S. patent application number 15/674175 was filed with the patent office on 2018-04-05 for analog-to-digital converter non-linearity correction using coefficient transformation.
The applicant listed for this patent is TEXAS INSTRUMENTS INCORPORATED. Invention is credited to Chandrasekhar SRIRAM, Jawaharlal TANGUDU.
Application Number | 20180097525 15/674175 |
Document ID | / |
Family ID | 61711548 |
Filed Date | 2018-04-05 |
United States Patent
Application |
20180097525 |
Kind Code |
A1 |
TANGUDU; Jawaharlal ; et
al. |
April 5, 2018 |
ANALOG-TO-DIGITAL CONVERTER NON-LINEARITY CORRECTION USING
COEFFICIENT TRANSFORMATION
Abstract
Circuitry for correcting non-linearity of an analog-to-digital
converter. A non-linearity correction system for an
analog-to-digital converter (ADC) includes coefficient storage,
coefficient transformation circuitry, and correction circuitry. The
coefficient storage is encoded with a first set of coefficients for
correcting non-linearity of the ADC at a first sampling rate. The
coefficient transformation circuitry is coupled to the coefficient
storage. The coefficient transformation circuitry is configured to
generate a second set of coefficients for correcting non-linearity
of the ADC at a different sampling rate. The correction circuitry
is configured to apply the second set of coefficients to correct
non-linearity in output of the ADC while the ADC is operating at
the different sampling rate.
Inventors: |
TANGUDU; Jawaharlal;
(Bangalore, IN) ; SRIRAM; Chandrasekhar; (Chennai,
IN) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
TEXAS INSTRUMENTS INCORPORATED |
Dallas |
TX |
US |
|
|
Family ID: |
61711548 |
Appl. No.: |
15/674175 |
Filed: |
August 10, 2017 |
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
H03M 1/0854 20130101;
H03M 1/08 20130101; H03M 1/06 20130101; H03M 1/1042 20130101; H03M
1/0836 20130101 |
International
Class: |
H03M 1/08 20060101
H03M001/08 |
Foreign Application Data
Date |
Code |
Application Number |
Oct 5, 2016 |
IN |
201641034098 |
Claims
1. (canceled)
2. A non-linearity correction system for an analog-to-digital
converter (ADC), the system comprising: coefficient storage encoded
with a first set of coefficients for correcting non-linearity of
the ADC at a first sampling rate; coefficient transformation
circuitry coupled to the coefficient storage, the coefficient
transformation circuitry configured to generate a second set of
coefficients for correcting non-linearity of the ADC at a different
sampling rate; and correction circuitry configured to apply the
second set of coefficients to correct non-linearity in an output of
the ADC while the ADC is operating at the different sampling rate;
wherein: the first set of coefficients is for correcting the
non-linearity of the ADC for a predetermined band of signal; and
the coefficient transformation circuitry is configured to generate
the second set of coefficients for a different band of the
signal.
3. The system of claim 2, wherein the different band is narrower
than and within the predetermined band, and the coefficient
transformation circuitry is configured to generate the second set
of coefficients for fewer than a number of terms corresponding to
the first set of coefficients based on the second band being
narrower than and within the predetermined band.
4. A non-linearity correction system for an analog-to-digital
converter (ADC), the system comprising: coefficient storage encoded
with a first set of coefficients for correcting non-linearity of
the ADC at a first sampling rate; coefficient transformation
circuitry coupled to the coefficient storage, the coefficient
transformation circuitry configured to generate a second set of
coefficients for correcting non-linearity of the ADC at a different
sampling rate; and correction circuitry configured to apply the
second set of coefficients to correct non-linearity in an output of
the ADC while the ADC is operating at the different sampling rate;
wherein the coefficient transformation circuitry is configured to
multiply the first set of coefficients by a transformation matrix
to generate the second set of coefficients.
5. A non-linearity correction system for an analog-to-digital
converter (ADC), the system comprising: coefficient storage encoded
with a first set of coefficients for correcting non-linearity of
the ADC at a first sampling rate; coefficient transformation
circuitry coupled to the coefficient storage, the coefficient
transformation circuitry configured to generate a second set of
coefficients for correcting non-linearity of the ADC at a different
sampling rate; and correction circuitry configured to apply the
second set of coefficients to correct non-linearity in an output of
the ADC while the ADC is operating at the different sampling rate;
wherein the first set of coefficients is selected for a virtual
sampling rate that is at least twice the maximum frequency
convertible by the ADC.
6. A non-linearity correction system for an analog-to-digital
converter (ADC), the system comprising: coefficient storage encoded
with a first set of coefficients for correcting non-linearity of
the ADC at a first sampling rate; coefficient transformation
circuitry coupled to the coefficient storage, the coefficient
transformation circuitry configured to generate a second set of
coefficients for correcting non-linearity of the ADC at a different
sampling rate; and correction circuitry configured to apply the
second set of coefficients to correct non-linearity in an output of
the ADC while the ADC is operating at the different sampling rate;
wherein the first set of coefficients is selected for use across
each of a plurality of Nyquist bands supported by the ADC.
7. The system of claim 6, wherein the coefficient transformation
circuitry is configured to generate the second set of coefficients
for fewer than a number of terms corresponding to the first set of
coefficients.
8. A non-linearity correction system for an analog-to-digital
converter (ADC), the system comprising: coefficient storage encoded
with a first set of coefficients for correcting non-linearity of
the ADC at a first sampling rate; coefficient transformation
circuitry coupled to the coefficient storage, the coefficient
transformation circuitry configured to generate a second set of
coefficients for correcting non-linearity of the ADC at a different
sampling rate; and correction circuitry configured to apply the
second set of coefficients to correct non-linearity in an output of
the ADC while the ADC is operating at the different sampling rate;
further comprising control circuitry coupled to the coefficient
transformation circuitry, the control circuitry configured to
communicate to the coefficient transformation circuitry one or more
parameters selected from the different sampling rate, a number of
terms corresponding to the second set of coefficients, or a band of
signal to be processed using the second set of coefficients.
9. The system of claim 8, wherein the coefficient transformation
circuitry is configured to: generate a transformation matrix based
on the one or more parameters; and multiply the first set of
coefficients by the transformation matrix to generate the second
set of coefficients.
10. (canceled)
11. Analog-to-digital conversion circuitry, comprising: an
analog-to-digital converter (ADC); and a non-linearity correction
circuit coupled to an output of the ADC, the non-linearity
correction circuit comprising: coefficient storage encoded with a
first set of coefficients for correcting non-linearity of the ADC
at a first sampling rate; coefficient transformation circuitry
coupled to the coefficient storage, the coefficient transformation
circuitry configured to: retrieve the first set of coefficients
from the coefficient storage; and process the first set of
coefficients to generate a second set of coefficients for
correcting non-linearity of the ADC at a second sampling rate; and
correction circuitry configured to apply the second set of
coefficients to correct non-linearity in output of the ADC while
the ADC is operating at the second sampling rate; wherein: the
first set of coefficients is for correcting the non-linearity of
the ADC for a predetermined band of signal; and the coefficient
transformation circuitry is configured to generate the second set
of coefficients for a different band of the signal with the band
being within the predetermined band.
12. The analog-to-digital conversion circuitry of claim 11, wherein
the different band is narrower than and within the predetermined
bandwidth, and the coefficient transformation circuitry is
configured to generate the second set of coefficients for fewer
than a number of terms corresponding to the first set of
coefficients based on the second band being narrower than and
within the predetermined band.
13. Analog-to-digital conversion circuitry, comprising: an
analog-to-digital converter (ADC); and a non-linearity correction
circuit coupled to an output of the ADC, the non-linearity
correction circuit comprising: coefficient storage encoded with a
first set of coefficients for correcting non-linearity of the ADC
at a first sampling rate; coefficient transformation circuitry
coupled to the coefficient storage, the coefficient transformation
circuitry configured to: retrieve the first set of coefficients
from the coefficient storage; and process the first set of
coefficients to generate a second set of coefficients for
correcting non-linearity of the ADC at a second sampling rate; and
correction circuitry configured to apply the second set of
coefficients to correct non-linearity in output of the ADC while
the ADC is operating at the second sampling rate; wherein the
coefficient transformation circuitry is configured to multiply the
first set of coefficients by a transformation matrix to generate
the second set of coefficients.
14. Analog-to-digital conversion circuitry, comprising: an
analog-to-digital converter (ADC); and a non-linearity correction
circuit coupled to an output of the ADC, the non-linearity
correction circuit comprising: coefficient storage encoded with a
first set of coefficients for correcting non-linearity of the ADC
at a first sampling rate; coefficient transformation circuitry
coupled to the coefficient storage, the coefficient transformation
circuitry configured to: retrieve the first set of coefficients
from the coefficient storage; and process the first set of
coefficients to generate a second set of coefficients for
correcting non-linearity of the ADC at a second sampling rate; and
correction circuitry configured to apply the second set of
coefficients to correct non-linearity in output of the ADC while
the ADC is operating at the second sampling rate; wherein the first
set of coefficients is selected for a virtual sampling rate that is
at least twice the maximum sampling rate of the ADC.
15. Analog-to-digital conversion circuitry, comprising: an
analog-to-digital converter (ADC); and a non-linearity correction
circuit coupled to an output of the ADC, the non-linearity
correction circuit comprising: coefficient storage encoded with a
first set of coefficients for correcting non-linearity of the ADC
at a first sampling rate; coefficient transformation circuitry
coupled to the coefficient storage, the coefficient transformation
circuitry configured to: retrieve the first set of coefficients
from the coefficient storage; and process the first set of
coefficients to generate a second set of coefficients for
correcting non-linearity of the ADC at a second sampling rate; and
correction circuitry configured to apply the second set of
coefficients to correct non-linearity in output of the ADC while
the ADC is operating at the second sampling rate; wherein the first
set of coefficients is selected for use across a plurality of
Nyquist bands supported by the ADC.
16. Analog-to-digital conversion circuitry, comprising: an
analog-to-digital converter (ADC); and a non-linearity correction
circuit coupled to an output of the ADC, the non-linearity
correction circuit comprising: coefficient storage encoded with a
first set of coefficients for correcting non-linearity of the ADC
at a first sampling rate; coefficient transformation circuitry
coupled to the coefficient storage, the coefficient transformation
circuitry configured to: retrieve the first set of coefficients
from the coefficient storage; and process the first set of
coefficients to generate a second set of coefficients for
correcting non-linearity of the ADC at a second sampling rate; and
correction circuitry configured to apply the second set of
coefficients to correct non-linearity in output of the ADC while
the ADC is operating at the second sampling rate; wherein the
coefficient transformation circuitry is configured to generate the
second set of coefficients for fewer than a number of terms
corresponding to the first set of coefficients.
17. Analog-to-digital conversion circuitry, comprising: an
analog-to-digital converter (ADC); and a non-linearity correction
circuit coupled to an output of the ADC, the non-linearity
correction circuit comprising: coefficient storage encoded with a
first set of coefficients for correcting non-linearity of the ADC
at a first sampling rate; coefficient transformation circuitry
coupled to the coefficient storage, the coefficient transformation
circuitry configured to: retrieve the first set of coefficients
from the coefficient storage; and process the first set of
coefficients to generate a second set of coefficients for
correcting non-linearity of the ADC at a second sampling rate; and
correction circuitry configured to apply the second set of
coefficients to correct non-linearity in output of the ADC while
the ADC is operating at the second sampling rate; further
comprising control circuitry coupled to the coefficient
transformation circuitry, the control circuitry configured to
communicate to the coefficient transformation circuitry one or more
parameters selected from the different sampling rate, a number of
terms corresponding to the second set of coefficients, or a band of
signal to be processed using the second set of coefficients.
18. The system of claim 17, wherein the coefficient transformation
circuitry is configured to: generate a transformation matrix based
on the one or more parameters; and multiply the first set of
coefficients by the transformation matrix to generate the second
set of coefficients.
19. A circuit for correcting non-linearity of an analog-to-digital
converter (ADC), the circuit comprising: coefficient storage
encoded with a first set of coefficients for correcting
non-linearity of the ADC at a first sampling rate and for a first
band of ADC output signal; coefficient transformation circuitry
coupled to the coefficient storage, the coefficient transformation
circuitry configured to multiply the first set of coefficients by a
transformation matrix to generate a second set of coefficients for
correcting non-linearity of the ADC at a second sampling rate and
for a second band of ADC output signal; and correction circuitry
configured to apply the second set of coefficients to correct
non-linearity in an output of the ADC while the ADC is operating at
the second sampling rate and with the second band of ADC output
signal.
20. The circuit of claim 19, wherein the second band is narrower
than and within the first band, and the coefficient transformation
circuitry is configured to generate the second set of coefficients
with fewer than a number of terms corresponding to the first set of
coefficients based on the second band being narrower than and
within the first band.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] The present application claims priority to Indian
Provisional Patent Application No. 201641034098, filed Oct. 5,
2016, titled "Non Linearity Corrector Coefficient Transformation,"
which is hereby incorporated herein by reference in its
entirety.
BACKGROUND
[0002] An analog-to-digital converter (ADC) is a circuit that
converts a continuous time signal (e.g., a voltage or current) into
a discrete time representation that is proportional in magnitude to
the continuous time signal ADCs are used for signal measurement in
a variety of electronics systems, such as mobile phones, audio and
video equipment, wireline communications systems, measurement
equipment, and radar systems. Various error sources affect ADC
performance. For example, ADCs of all types may suffer from
quantization error, non-linearity, and/or clock jitter.
SUMMARY
[0003] Circuitry for correcting non-linearity of an
analog-to-digital converter (ADC) is disclosed herein. In one
embodiment, a non-linearity correction system for an ADC includes
coefficient storage, coefficient transformation circuitry, and
correction circuitry. The coefficient storage is encoded with a
first set of coefficients for correcting non-linearity of the ADC
at a first sampling rate. The coefficient transformation circuitry
is coupled to the coefficient storage. The coefficient
transformation circuitry is configured to generate a second set of
coefficients for correcting non-linearity of the ADC at a different
sampling rate. The correction circuitry is configured to apply the
second set of coefficients to correct non-linearity in output of
the ADC while the ADC is operating at the different sampling
rate.
[0004] In another embodiment, analog-to-digital conversion
circuitry includes an ADC and a non-linearity correction circuit
coupled to an output of the ADC. The non-linearity correction
circuit includes coefficient storage, coefficient transformation
circuitry, and correction circuitry. The coefficient storage is
encoded with a first set of coefficients for correcting
non-linearity of the ADC at a first sampling rate. The coefficient
transformation circuitry is coupled to the coefficient storage. The
coefficient transformation circuitry is configured to retrieve the
first set of coefficients from the coefficient storage, and to
process the first set of coefficients to generate a second set of
coefficients for correcting non-linearity of the ADC at a second
sampling rate. The correction circuitry is configured to apply the
second set of coefficients to correct non-linearity in output of
the ADC while the ADC is operating at the second sampling rate.
[0005] In a further embodiment, a circuit for correcting
non-linearity of an ADC includes coefficient storage, coefficient
transformation circuitry, and correction circuitry. The coefficient
storage encoded with a first set of coefficients for correcting
non-linearity of the ADC at a first sampling rate and for a first
bandwidth of ADC output signal. The coefficient transformation
circuitry is coupled to the coefficient storage. The coefficient
transformation circuitry is configured to multiply the first set of
coefficients by a transformation matrix to generate a second set of
coefficients for correcting non-linearity of the ADC at a second
sampling rate and for a second bandwidth of ADC output signal. The
correction circuitry is configured to apply the second set of
coefficients to correct non-linearity in output of the ADC while
the ADC is operating at the second sampling rate and with the
second bandwidth of ADC input signal.
BRIEF DESCRIPTION OF THE DRAWINGS
[0006] For a detailed description of various examples, reference
will now be made to the accompanying drawings in which:
[0007] FIG. 1 shows a block diagram of a digitization system that
includes digital non-linearity correction in accordance with
various embodiments;
[0008] FIG. 2 shows a block diagram of a non-linearity correction
system in accordance with various embodiments;
[0009] FIG. 3 shows coefficient transformation circuitry in
accordance with various embodiments;
[0010] FIG. 4 shows composition of a transformation matrix and
application of the transformation matrix to generate non-linearity
correction coefficients in accordance with various embodiments;
[0011] FIG. 5 shows a band of interest spanning multiple Nyquist
zones with respect to a sampling rate (f.sub.s0) corresponding to
stored non-linearity coefficients;
[0012] FIGS. 6 and 7 show coefficient transformation circuitry in
accordance with various embodiments;
[0013] FIG. 8 shows third harmonic distortion performance using a
non-linearity correction system in accordance with various
embodiments; and
[0014] FIG. 9 shows a flow diagram for a method for non-linearity
correction in accordance with various embodiments.
DETAILED DESCRIPTION
[0015] Certain terms are used throughout the following description
and claims to refer to particular system components. As one skilled
in the art will appreciate, different companies may refer to a
component by different names. This document does not intend to
distinguish between components that differ in name but not
function. In the following discussion and in the claims, the terms
"including" and "comprising" are used in an open-ended fashion, and
thus should be interpreted to mean "including, but not limited to .
. . ." Also, the term "couple" or "couples" is intended to mean
either an indirect or direct wired or wireless connection. Thus, if
a first device couples to a second device, that connection may be
through a direct connection or through an indirect connection via
other devices and connections. The recitation "based on" is
intended to mean "based at least in part on." Therefore, if X is
based on Y, X may be a function of Y and any number of additional
factors.
[0016] In some applications, analog-to-digital converters (ADCs)
must meet stringent harmonic distortion and/or intermodulation
distortion specifications. For example, ADCs used in wireless
infrastructure must meet strict third harmonic distortion and
intermodulation distortion requirements to conform to cellular
system standards. Harmonic distortion (e.g., third harmonic
distortion) and intermodulation distortion of an ADC can be
improved by applying a factory calibrated digital non-linearity
correction to the output of the ADC. Because an ADC may operate
over a wide range of sampling rates, and good non-linearity
performance must be provided for all sampling rates, in some
non-linearity correction systems, non-linearity correction
coefficients may be provided for each supported sampling rate, and
stored in memory. Unfortunately, the amount of memory required to
store the coefficients increases linearly with the number of
sampling rates. Consequently, such non-linearity correction systems
may support only a few sampling rates.
[0017] Embodiments of the non-linearity correction system disclosed
herein store non-linearity coefficients for one sampling rate and
transform the coefficients for use with any sampling rate at which
an ADC operates. The non-linearity correction system may apply
matrix multiplication to transform the non-linearity coefficients.
Accordingly, the memory dedicated to storage of non-linearity
coefficients may be reduced and the number of sampling rates
supported is not limited by the number of stored coefficient sets.
Furthermore, because fewer coefficients are required for
non-linearity correction in a narrow band than are required in a
wide band, embodiments may reduce the number of coefficients
applied in non-linearity correction as a function of bandwidth, and
reduce the power consumed by non-linearity correction.
[0018] FIG. 1 shows a block diagram of a digitization system 100
that includes digital non-linearity correction in accordance with
various embodiments. The system 100 includes an ADC 106, a
non-linearity correction system 102 and control circuitry 104. The
ADC 106 receives an analog signal as input and converts the analog
signal to a stream of digital values (digital samples) 108. The ADC
106 may be a direct-conversion ADC, successive approximation ADC,
or any other type of ADC. The ADC 106 may support multiple sampling
frequencies (i.e., sampling rates). The digital samples produced by
the ADC 106 are provided to the non-linearity correction system
102.
[0019] The non-linearity correction system 102 processes the
digital samples 108 received from the ADC 106 to correct the
digital samples for non-linearity of the ADC 106. Some embodiments
of the non-linearity correction system 102 may apply a third order
non-linearity correction model using both Volterra terms and
derivative signal as:
y ( n ) = x ( n - L 2 ) + y 3 ( n ) + d 3 ( n ) x d ( n ) = x ' ( n
) Derivative Signal y 3 ( n ) = k 1 = 0 L k 2 = k 1 L k 3 = k 2 L c
3 ( k 1 , k 2 , k 3 ) x ( n - k 1 ) x ( n - k 2 ) x ( n - k 3 ) d 3
( n ) = k 1 = 0 L c 3 , 1 ( k 1 ) x 2 ( n - k 1 ) x d ( n - L 2 ) +
k 1 = 0 L c 3 , 2 ( k 1 ) x ( n - k 1 ) x d ( n - L 2 ) 2 + c 3 , 3
x d ( n - L 2 ) 3 ( 1 ) ##EQU00001##
where: [0020] x(n) is the input to the non-linearity correction
system 102 (i.e., the output of the ADC 106; [0021] y(n) is the
output of the non-linearity correction system 102 (i.e.,
non-linearity corrected samples); [0022] L is memory order; and
[0023] k.sub.1, k.sub.2, and k.sub.3 are temporary index
variables.
[0024] In the correction equations above, the main input signal
x(n) is added with 3.sup.rd order non-linearity correction terms
y.sub.3(n) and d.sub.3(n). y.sub.3(n) is a standard Volterra model
based term and d.sub.3(n) has 3.sup.rd order terms using both the
derivative signal and main signal.
c.sub.3(k.sub.1,k.sub.2,k.sub.3), c.sub.3,1(k.sub.1),
c.sub.3,2(k.sub.1)c.sub.3,3, are coefficients in this non-linearity
correction model. The values of the coefficients are programmed
based on the ADC sampling rate. Similar equations can be written
for second order, fourth order or any other order non-linearity
terms, and may use standard Volterra terms and other non-linearity
terms.
[0025] Embodiments of the non-linearity correction system 102 may
apply more or fewer terms than are shown in equation (1) as needed
to correct non-linearity of a given set of digital samples 108. The
coefficients applied to correct non-linearity in the digital
samples 108 may be derived, by the non-linearity correction system
102, from a single set of non-linearity coefficients stored in the
non-linearity correction system 102. The non-linearity correction
system 102 may derive non-linearity coefficients for any sampling
rate supported by the ADC 106 from the single set of non-linearity
coefficients.
[0026] The control circuitry 104 generates control signals and
information that the non-linearity correction system 102 applies to
generate the coefficients applied to the digital samples 108 from
the single set of coefficients stored in the non-linearity
correction system 102. For example, the control circuitry 104 may
communicate the sampling rate at which the ADC 106 is operating,
information concerning the bandwidth of signal to be corrected for
non-linearity, and other information to the non-linearity
correction system 102. The non-linearity correction system 102
applies the information to generate the coefficients applied to
reduce non-linearity in the digital samples 108.
[0027] Some embodiments of the system 100 may include multiple
instances of the ADC 106, where each ADC 106 has different
non-linearity performance. Because the coefficients applied to
correct non-linearity in the digital samples generated by the ADCs
106 may be derived, by the non-linearity correction system 102,
from a single set of non-linearity coefficients stored in the
non-linearity correction system 102, the storage requirements of
the non-linearity system 102 may be reduced relative to a system
that stores non-linearity coefficients for each ADC 106 at each
sampling frequency.
[0028] FIG. 2 shows a block diagram of the non-linearity correction
system 102 in accordance with various embodiments. The
non-linearity correction system 102 includes coefficient storage
202, coefficient transformation circuitry 204, and non-linearity
correction circuitry 206. The coefficient storage 202 stores a
single set of non-linearity correction coefficients 208. The
coefficients 208 may be directly applicable to only a single
sampling rate of the ADC 106. The coefficient storage 202 may
provide storage for the coefficients 208 using any of a variety of
data storage or memory technologies. For example, the coefficient
storage 202 may include fusible link technology, anti-fuse
technology, FLASH memory cell technology, or other data storage
technology suitable to record the coefficients 208. The coefficient
storage 202 is coupled to the coefficient transformation circuitry
204.
[0029] The coefficient transformation circuitry 204 retrieves the
coefficients 208 from coefficient storage 202 and applies a
transform to the coefficients 208 to produce the coefficients
needed to correct non-linearity in the digital samples 108 received
from the ADC 106. The coefficient transformation circuitry 204 may
multiply the coefficients 208 by a transformation matrix to produce
the coefficients applied to the digital samples 108. Parameters
used to generate the transformation matrix may be provided to the
coefficient transformation circuitry 204 by the control circuitry
104 as TRANSFORM CONTROL. The coefficient transformation circuitry
204 provides the transformed coefficients to the non-linear
correction circuitry 206. Some embodiments of the coefficient
transformation circuitry 204 may include a processor, such as a
general-purpose microprocessor, a digital signal processor, a
microcontroller, or other instruction execution device, and
instructions executable by the processor to generate the
transformation matrix and apply the transformation matrix to
generate the coefficients needed to correct non-linearity in the
digital samples received from the ADC 106 from the coefficients
208. Some embodiments of the coefficient transformation circuitry
204 may be implemented using dedicated hardware circuitry (e.g.,
dedicated digital circuitry).
[0030] The non-linearity correction circuitry 206 applies the
transformed coefficients received from the coefficient
transformation circuitry 204 to correct non-linearity in the
digital samples 108. Embodiments of the non-linearity correction
circuitry 206 may apply the transformed coefficients as per
equation (1) to produce non-linearity corrected samples. Some
embodiments of the non-linearity correction circuitry 206 may
include a processor, such as a general-purpose microprocessor, a
digital signal processor, a microcontroller, or other instruction
execution device, and instructions executable by the processor to
apply the coefficients received from the coefficient transformation
circuitry 204 to the digital samples 108. Some embodiments of the
non-linearity correction circuitry 206 may be implemented using
dedicated hardware circuitry (e.g., dedicated digital
circuitry).
[0031] In some embodiments of the non-linearity correction system
102, the non-linearity correction circuitry 206, the coefficient
transformation circuitry 204, and the coefficient storage 202 are
integrated on a single device. In some embodiments, the
non-linearity correction circuitry 206, the coefficient
transformation circuitry 204, and the coefficient storage 202 may
be provided two or more different devices. For example, the
coefficient transformation circuitry 204 may be disposed on a first
device and non-linearity coefficients transferred from the first
device to a second device that contains the non-linearity
correction circuitry 206 for use in correcting the output of the
ADC 106.
[0032] FIG. 3 shows coefficient transformation circuitry 204 in
accordance with various embodiments. The coefficient transformation
circuitry 204 retrieves the coefficients 208 from coefficient
storage 202 and applies a transform to the coefficients 208 to
produce the coefficients needed to correct non-linearity in the
digital samples 108 received from the ADC 106. The coefficients 208
correspond to an ADC sampling frequency f.sub.s0. Transform control
information received from the control circuitry 104 includes
f.sub.s0, the operating sampling rate (f.sub.s) of the ADC 106, the
frequency band (BAND) of the digital samples 108 to be
non-linearity corrected, and a number non-linearity terms (NL TERMS
or terms) to be applied to the digital samples 108. The coefficient
transformation circuitry 204 determines the coefficients at
sampling rate f.sub.s from the coefficients at the sampling rate
f.sub.s0. The transformation assumes that non-linearity behavior of
the ADC 106 is independent of sampling rate. The coefficient
transformation circuitry 204 computes the transformed coefficients
as:
C(f.sub.s)=(C(f.sub.s0), f.sub.s, f.sub.s0, f.sub.1, f.sub.h,
terms) (2)
where: [0033] is the coefficient transformation function; [0034]
f.sub.1 and f.sub.h are respectively the lower and higher edge
frequencies of the band of interest; and [0035] C is the column
vector of non-linearity coefficients 108 (which are the weights of
the NL terms) as a function of the sampling rate.
[0036] Representing the coefficient transformation as a matrix
multiplication:
C(f.sub.s)=TC(f.sub.s0) (3)
where T is the transformation matrix, embodiments formulate the
level of the non-linearity spurs as a linear combination of the
non-linearity coefficients. The weights of the coefficients change
with sampling rate, but results of the linear combination that is
the spur level are constant. Where H.sub.0 and H.sub.1 are each
matrices representative of non-linearity response of the ADC 106 at
a sampling frequency, given a linear combination (H.sub.0) of the
non-linearity coefficients at sampling frequency f.sub.s0 and a
linear combination (H.sub.1) of the non-linearity coefficients at
sampling frequency f.sub.s, the transformation matrix T is computed
as the product of H.sub.0 and the pseudoinverse of H.sub.1:
T=pinv(H.sub.1)H.sub.0 (4)
[0037] In some embodiments of the non-linearity correction system
102, the transformation matrices applied to generate non-linearity
coefficients are stored in the coefficient transformation circuitry
204 or memory accessed by the coefficient transformation circuitry
204. In some embodiments of the non-linearity correction system
102, the non-linearity response matrices (H.sub.x) are derived by
the coefficient transformation circuitry 204 and the transformation
matrices are computed as per equation (4) and further described
below.
[0038] Generation of the non-linearity response matrices H.sub.0
and H.sub.1, and the transformation matrix T is now described in
further detail. When tones are input to a non-linear device, such
as the ADC 106, non-linearity spurs are produced. For example, when
2 tones at f.sub.1 and f.sub.2 are input to a non-linear device
having third order non-linearity spurs are produced at 3f.sub.1,
3f.sub.2, 2f.sub.1-f.sub.2, 2f.sub.2-f.sub.1, 2f.sub.1+f.sub.2, and
2f.sub.2+f.sub.1. The level and phase of these spurs relative to
the level and phase of input tones determine the non-linearity
response.
[0039] If the ADC 106 operating at f.sub.s0, for a tone at
frequency f in the band of interest, normalized (.omega.) is
defined as:
.omega.=2.pi.f/f.sub.s (5)
[0040] As an example, using a third non-linearity model comprising
of 3 terms, namely x.sup.3(n), x(n)x(n-1)x(n-2), and
x.sup.2(n)x'(n), then the non-linearity response at 3.omega. when a
tone at .omega. with amplitude A is input can be written as:
NLResp@3f=A.sup.3[c.sub.1+c.sub.2e.sup.-j3.omega.+c.sub.3j.omega.]
(6)
where c.sub.1, c.sub.2, and c.sub.3 are the coefficients of the
terms x.sup.3(n), x(n)x(n-1)x(n-2), and x.sup.2(n)x'(n)
respectively.
[0041] Similarly, when 2 tones and 3 tones are input, the
non-linearity response (at each non-linearity spur) can be written
as a linear combination of the coefficients (c.sub.1, c.sub.2,
c.sub.3) where the weights of the coefficients are a function of
the non-linearity terms being used and the sampling frequency
(f.sub.s). Thus, for tones covering the band of interest, a system
of linear equations of the coefficients is generated. In matrix
form this system of equations can be written as:
H{right arrow over (C)}={right arrow over (R)} (7)
where {right arrow over (C)} is a vector of coefficients, {right
arrow over (R)} is a vector of the non-linearity response, and the
weight matrix (H) is a function of the sampling rate for a given
non-linearity model.
[0042] For 2 different sampling rates f.sub.s0 and f.sub.s1, to
find the coefficients at f.sub.s1 given coefficients at f.sub.s0
the non-linearity response is equated,
H.sub.1{right arrow over (C)}.sub.1=H.sub.0{right arrow over
(C)}.sub.0 (8)
{right arrow over (C)}.sub.1=pinv(H.sub.1)H.sub.0{right arrow over
(C)}.sub.0 (9)
Define T=pinv(H.sub.1)H.sub.0 (10)
where, C.sub.0, C.sub.1 are the coefficient vectors at f.sub.s0 and
f.sub.s1 respectively and H.sub.0, H.sub.1 are the weight matrices
at f.sub.s0 and f.sub.s1 respectively.
[0043] The number of rows in H.sub.0 is equal to the number of
equations where each equation corresponds to a specific
non-linearity response. The number of columns is equal to the
number of coefficients. The number of rows should typically be the
same or greater than the number of columns. "T" is the
transformation matrix used to generate coefficients at sampling
rate f.sub.s1 from coefficients at sampling rate f.sub.s0. It is
expected that the non-linearity response covers the band of
interest by choosing a sufficient number of frequencies for single
tone, two tone, and three tones signal types. Embodiments of the
coefficient transformation circuitry 204 may derive the Hx matrices
as described and/or apply Hx matrices to produce a transformation
matrix (T). Systems external to the coefficient transformation
circuitry 204 may derive the Hx matrices and/or the transformation
matrix (T) and provide the Hx matrices or the transformation matrix
(T) to the coefficient transformation circuitry 204 for use in
generating non-linearity coefficients.
[0044] FIG. 4 shows application of the transformation matrix T in
the coefficient transformation circuitry 204 to produce the
coefficients for correcting non-linearity in digital samples 108
produced by the ADC 106 at sampling frequency fs from coefficients
208 for digital samples at sampling frequency f.sub.s0.
[0045] The ADC 106 may produce digital samples 108 corresponding
one of multiple Nyquist zones. For example, with the ADC 106
undersampling may be used to reflect data from a higher Nyquist
zone into a lower Nyquist zone. Given a set of non-linearity
coefficients 208 corresponding to a sampling frequency f.sub.s0,
the Nyquist zones relevant to the coefficients may not coincide
with the operational Nyquist zones of the ADC 106. FIG. 5 shows
multiple Nyquist zones of the ADC 106 and a band of interest with
multiple overlapping Nyquist zones corresponding to the sampling
frequency f.sub.s0. In FIG. 5, the band of interest is wholly
within a Nyquist zone of the ADC 106, but spans two Nyquist zones
with respect to f.sub.s0. As a result coefficient transformation
using coefficients 208 corresponding to sampling frequency f.sub.s0
is problematic. As a result of overlapping Nyquist zones
corresponding to sampling frequency f.sub.s0, NL coefficients for
this band cannot be represented.
[0046] To overcome the issue of a band of interest overlapping two
Nyquist zones, embodiments of the non-linearity correction system
102 may provide the coefficients 208 as representing non-linearity
response over all possible frequency bands supported by the ADC 106
at a virtual sampling rate f.sub.sv where f.sub.sv>2B and B is
the maximum frequency supported by the ADC 106. Such embodiments of
the non-linearity correction system 102 use a larger number of
non-linearity terms to represent the complete non-linearity
response, however, non-linearity correction may use only a required
lesser number of non-linearity terms.
[0047] FIG. 6 shows coefficient transformation circuitry 204 in
accordance with various embodiments. The coefficient transformation
circuitry 204 retrieves the coefficients 208 from coefficient
storage 202 and applies a transform to the coefficients 208 to
produce the coefficients needed to correct non-linearity in the
digital samples received from the ADC 106. The coefficients 208
correspond to a virtual sampling frequency f.sub.sv. Transform
control information received from the control circuitry 104
includes f.sub.sv, the operating sampling rate (f.sub.s), the
frequency band (BAND) of the digital samples 108 to be
non-linearity corrected, and a set of non-linearity terms at the
virtual sampling frequency (NL TERMS AT f.sub.sv or termsfsv), and
a set of non-linearity terms at the operating sampling frequency
(NL TERMS AT f.sub.s or termsfs). The number of non-linearity terms
at the virtual sampling frequency may be greater than the number of
non-linearity terms at the operating sampling frequency. The
coefficient transformation circuitry 204 determines the
coefficients at sampling rate f.sub.s from the coefficients at the
sampling rate f.sub.sv. The coefficient transformation circuitry
204 computes the transformed coefficients as:
C(f.sub.s)=(C(f.sub.sv), f.sub.s, f.sub.sv, f.sub.1, f.sub.h,
terms.sub.f.sub.s, terms.sub.f.sub.sv) (11)
where: [0048] is the coefficient transformation function; [0049]
f.sub.1 and f.sub.h are respectively the lower and higher edge
frequencies of the band of interest; and [0050] C is the column
vector of non-linearity coefficients (which are the weights of the
NL terms) as a function of the sampling rate. The length of the
column vector is the number of terms and hence could be different
for f.sub.sv and f.sub.s.
[0051] In some applications, the signal conversion bandwidth of the
ADC 106 may vary. For example, the signal conversion bandwidth of
the ADC 106 may be wide or narrow in radio frequency applications.
To support wideband use, the coefficients 208 may support the
entire width of a Nyquist zone. Wideband non-linearity correction
requires a larger number of non-linearity terms than narrowband
non-linearity correction. Embodiments of the non-linearity
correction system 102 may apply fewer terms to correct
non-linearity in narrowband data than are applied to correct
non-linearity in wideband data. Accordingly, the coefficient
transformation circuitry 204 may produce fewer non-linearity
coefficients, via transformation, for correction of narrowband data
than are produced for wideband data. Reducing the number of terms
applied by the non-linear correction circuitry 206 reduces the
overall power consumption of the non-linearity correction system
102.
[0052] FIG. 7 shows coefficient transformation circuitry 204 in
accordance with various embodiments. The coefficient transformation
circuitry 204 retrieves the coefficients 208 from coefficient
storage 202 and applies a transform to the coefficients 208 to
produce the coefficients needed to correct non-linearity in the
digital samples received from the ADC 106. The coefficients 208 are
for wideband data and correspond to the band encompassed by an
entire Nyquist zone. Transform control information received from
the control circuitry 104 includes the operating sampling rate
(f.sub.s), the frequency band (BAND) of the digital samples 108 to
be non-linearity corrected, a number non-linearity terms for
wideband data (NL TERMS FOR WIDEBAND or termsWb), and a number of
non-linearity terms for narrowband data (NL TERMS FOR NARROWBAND or
termsNB). The number of non-linearity terms for wideband data may
be greater than the number of non-linearity terms for narrowband
data. The coefficient transformation circuitry 204 determines the
coefficients for narrowband data at sampling rate f.sub.s as:
C(NB)=(C(WB), f.sub.s, f.sub.1, f.sub.h, terms.sub.WB,
terms.sub.NB) (12)
where: [0053] is the coefficient transformation function; [0054]
f.sub.1 and f.sub.h are respectively the lower and higher edge
frequencies of the band of interest; and [0055] C is the column
vector of non-linearity coefficients (which are the weights of the
corresponding NL terms) as a function of the sampling rate.
[0056] FIG. 8 shows third harmonic distortion performance using a
non-linearity correction system 102 that applies transformed
wideband coefficients to correct non-linearity in a narrow band as
described with respect to the coefficient transformation circuitry
204 of FIG. 7.
[0057] FIG. 9 shows a flow diagram for a method 900 for
non-linearity correction in accordance with various embodiments.
Though depicted sequentially as a matter of convenience, at least
some of the actions shown can be performed in a different order
and/or performed in parallel. Additionally, some embodiments may
perform only some of the actions shown. In some embodiments, at
least some of the operations of the method 900 can be performed by
the circuitry of the non-linearity correction system 102.
[0058] In block 902, the coefficient transformation circuitry 204
receives the coefficients 208 from the coefficient storage 202. The
coefficients 208 may be directed to correction of non-linearity in
digital samples 108 produced by the ADC 106 at a single given
sampling frequency. The digital samples 108 to be processed by the
method 900 are produced by the ADC 106 at a sampling frequency that
is different from the single given sampling frequency corresponding
to the coefficients 208. In some embodiments, the coefficients 208
may correspond to a virtual sampling frequency that is at least
twice the maximum frequency supported by the ADC 106. Such
embodiments include a large number of coefficients to represent the
complete non-linearity response. In some embodiments, the
coefficients 208 are for wideband data and correspond to the band
encompassed by an entire Nyquist zone of the ADC 106. Such
embodiments may also include a relatively large number of
coefficients.
[0059] In block 904, the coefficient transformation circuitry 204
generates a transformation matrix for derivation of the
coefficients to be applied to the digital samples 108 received from
the ADC 106. The coefficient transformation circuitry 204 may
generate the transformation matrix as described herein with respect
to equation (4).
[0060] In block 906, the coefficient transformation circuitry 204
performs a matrix multiplication of coefficients 208 and the
transformation matrix produced in block 904. The product of the
multiplication is transformed coefficients that can be applied to
correct non-linearity in the digital samples 108. The coefficient
transformation circuitry 204 may produce the transformed
coefficients to correct non-linearity in a band that is narrower
than and within the band represented by the coefficients 208. The
coefficient transformation circuitry 204 provides the transformed
coefficients to the non-linearity correction circuitry 206.
[0061] In block 908, the non-linearity correction circuitry 206
applies the transformed coefficients to correct non-linearity in
the digital samples 108 received from the ADC 106. In some
embodiments, the number of transformed coefficients applied to
correct non-linearity is a function of the bandwidth of interest in
the digital samples 108. Accordingly, the number of terms, and
corresponding number of coefficients, applied by the non-linearity
correction circuitry 206 may be reduced with bandwidth, with
commensurate reduction in the power consumed by the non-linearity
correction circuitry 206 to correct non-linearity in the digital
samples 108.
[0062] The above discussion is meant to be illustrative of the
principles and various embodiments of the present invention.
Numerous variations and modifications will become apparent to those
skilled in the art once the above disclosure is fully appreciated.
It is intended that the following claims be interpreted to embrace
all such variations and modifications.
* * * * *