Electricity: measuring and testing – Measuring – testing – or sensing electricity – per se – Analysis of complex waves
Reexamination Certificate
2000-12-14
2004-04-20
Nolan, Jr., Charles H. (Department: 2854)
Electricity: measuring and testing
Measuring, testing, or sensing electricity, per se
Analysis of complex waves
C324S076190, C324S076210
Reexamination Certificate
active
06724177
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to the measurement of communications signals.
2. State of the Art
Predistortion is commonly used in communications signal transmitters to compensate for signal distortion introduced by transmitter components. In wireless communications, for example, the final stage of a power amplifier is often the source of significant distortion, in particular AM/AM distortion and AM/PM distortion. Device characterization attempts to identify the distortion characteristics of a device to enable predistortion to be performed, the predistortion and the distortion ideally cancelling each other, leaving a distortion-free signal.
Device characterization requires signal measurement. Measurement circuitry, however, may also introduce significant distortion. In the case of a wireless communications system, for example, a typical characterization setup for characterizing a power amplifier is as shown in
FIG. 1. A
suitably-chosen RF stimulus (or “probe signal”)
101
is applied to a power amplifier
103
, which produces a communications signal
105
. The communications signal is applied to an I/Q demodulator and data converter
107
, which produces digital I and Q signals
109
. An I/Q compensator
111
alters the digital I and Q signals to compensate to the extent possible for non-idealities of the I/Q demodulator. Compensated digital I and Q
113
are input to a data analysis unit
115
(e.g., a stored-program computer), which produces estimates of an AM/AM distortion curve of the power amplifier and a AM/PM distortion curve of the power amplifier.
The amplifier of
FIG. 1
may be a conventional linear amplifier, as shown in
FIG. 2
, or may be a non-linear amplifier of a type shown in FIG.
3
.
A more detailed diagram of a main portion of the conventional measurement setup of
FIG. 1
is shown in FIG.
4
. An output signal
405
of a power amplifier
403
is applied to an RF port of a mixer
407
. Applied to a local oscillator port of the mixer is a local oscillator signal
409
having a frequency equal to the sum of a carrier frequency f
c
of the RF signal and a desired intermediate frequency f
IF
. The carrier frequency may be in the range of many hundreds of megahertz, for example, while the intermediate frequency may be in the range of many tens of megahertz. The mixer
407
produces an output signal
411
at the intermediate frequency. The intermediate frequency signal is split into two signals, one of which is applied to an I (in-phase processing) chain
420
I
and one of which is applied to a Q (quadrature processing) chain
420
Q
. The I and Q chains are largely identical, each having a mixer
421
, followed by a low-pass filter
423
, a video amplifier
425
, a further low-pass filter
427
, and an A/D converter
429
. A local oscillator (not shown) produces a reference signal
431
that is applied directly to the mixer of the I chain. A 90 degree phase shifter
433
produces a quadrature version of the reference signal, which is applied to the mixer of the Q chain. Digital outputs of the I and Q chains are combined to form a single complex number stream z(t), the combination being effected by a j operator
435
and a summation block
437
.
The foregoing setup is acceptable where accuracies of 10%, or perhaps even some few percent, are acceptable. In high-data-rate wireless communications systems, however, accuracies of less than 1% may be required. The conventional setup of
FIG. 1
does not allow for such accuracy to be obtained.
More particularly, the I/Q video amplifiers
423
are a source of various types of errors, including I/Q gain mismatch, different DC offsets, and non-linearity. The standard approach to signal measurement is particularly susceptible to the latter two sources of error. Furthermore, the power amplifier
403
itself exhibits what may also be termed DC offset, as a result of AM/AM distortion.
Assume that the RF stimulus
401
exhibits a sawtooth envelope. Whereas ideally the output signals of the I/Q video amplifiers would have a sawtooth envelope varying between a minimum amplitude a
0
and a maximum amplitude a
1
, as illustrated in
FIG. 5
, instead, these amplitude levels are shifted by a combination of DC offsets from the power amplifier (the offset of interest, caused by AM/AM distortion of the power amplifier, denoted herein by DC
PA(AM/AM)
) and from the I/Q video amplifiers (“nuisance” offsets, denoted herein by DC
VA
). Identifying the different offset components is analogous to being asked to guess two numbers given their sum—the offsets are, in practical terms, inseparable.
An additional source of significant error is the phase shifter
433
, which typically will have a tolerance of a few degrees of phase. If an accuracy of a fraction of a degree is desired, clearly such an error is intolerable.
SUMMARY OF THE INVENTION
The present invention, generally speaking, provides a method and apparatus for accurately measuring a communications signal. In accordance with one aspect of the invention, DC offset effects and nonlinearities attributable to a communications amplifier are made spectrally separable from DC offset effects and nonlinearities attributable to the measurement apparatus. Spectral separability may be accomplished, for example, by adding an offset frequency to a local oscillator used by a downconverter of the measurement apparatus. As a result, the signal of interest is moved away from baseband (zero frequency) to the offset frequency. Similarly, other nonlinearities in the video amplifier (such as third order distortion) manifest theselves mostly at harmonics of the offset frequency. Hence spectral separability between the power amplifier characteristics and these nonlinear impairments of the measurement system is achieved. At the offset frequency, the signal of interest may be represented by a single signal component (e.g., the in-phase component I) instead of requiring an I/Q representation. The measurement apparatus may therefore be dramatically simplified, and errors stemming from undesired interaction of I and Q signal processing chains may be avoided. In accordance with another aspect of the invention, the effects of noise are mitigated through different averaging methods.
REFERENCES:
patent: 2929989 (1960-03-01), Hurvitz
patent: 5789927 (1998-08-01), Belcher
patent: 5832247 (1998-11-01), Gildea et al.
patent: 6018784 (2000-01-01), Gildea et al.
patent: 6078290 (2000-06-01), McBurney et al.
patent: 6134223 (2000-10-01), Burke et al.
patent: 6161004 (2000-12-01), Galal et al.
patent: 6208844 (2001-03-01), Abdelgany
patent: 6211663 (2001-04-01), Moulthrop et al.
patent: 6243430 (2001-06-01), Mathe
patent: 6298103 (2001-10-01), Huang et al.
patent: 0499346 (1988-03-01), None
Ken Kundert, “Introduction to RF Simulation and Its Application” 0-7803-4497-Sep./1998/$10.00 © 1998 IEEE
Hamdan Waseem H.
Nolan, Jr. Charles H.
Thelen Reid & Priest LLP
Tropian Inc.
LandOfFree
Method and apparatus for accurate measurement of... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for accurate measurement of..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for accurate measurement of... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3217566