Method for self-calibrating a frequency of a modulator...

Details Method for self-calibrating a frequency of a modulator... Method for self-calibrating a frequency of a modulator...

2002-12-27

2004-11-23

JeanPierre, Peguy (Department: 2819)

Coded data generation or conversion

Analog to or from digital conversion

Differential encoder and/or decoder

C341S144000, C341S145000

Reexamination Certificate

active

06822592

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates to a method for self-calibrating a frequency of a modulator circuit, and circuit using said method, particularly but not exclusively a method for self-calibrating a frequency of center band of a band pass sigma-delta type modulator circuit.
BACKGROUND OF THE INVENTION
The sigma-delta modulators, with high performance characteristics, are used in a various number of applications, such as, by way of example, in audio communication systems.
Particularly, said applications require high range dynamic characteristic, for example, by means of a digital resolution of twelve bits, a precise center band frequency, having a value of ten Mhz, a small physical dimension of the modulator and a low current consumption.
Sigma-delta type band passing multi bit modulators are used to reach said performance and physical characteristics.
Referring to
FIG. 1
, a basic scheme of a sigma—delta modulator is shown, having a loop circuit
1
composed of an adding node
2
, a go path
3
and a return path
4
, also called feedback path.
In
FIG. 1
it is evident that the go path
3
is realized by the series of a first block
5
with a second block
6
, wherein said first block
5
is a filtering block and said second block
6
is an analog to digital conversion (ADC) block. The feedback path
4
is, instead, a digital to analog conversion (DAC) block.
The operation of the circuit scheme of
FIG. 1
is well known to a skilled person and therefore will not be shown.
In the modulator
1
, the filtering block
5
, in the easiest implementation, may be made, by way of example, according to an integrator filter scheme of the first order, as shown in
FIG. 2
, which is well known to an individual skilled in the art.
As shown in
FIG. 2
, the filtering block
5
is composed by a block
7
, having in the feedback path an amplifier
8
.
Particularly the block
7
is a discrete filter having as transfer function H(z) the following relationship:
H
(
z
)=
a
1
*z
−1
/(1
−b
1
*z
−1
)  (1)
while the amplifier
8
has an opportune amplifying value suitable for the embodiment of the sigma-delta modulator.
Referring now to
FIG. 3
, an output spectrum from a modulator having, as a filtering block, a filter of the third order is shown, including an abscissa axis indicating the frequency evaluated in MHz, and an ordinate axis indicating the output noise spectrum evaluated in dB, a first graph
9
, a second graph
10
and an observation band
12
.
The first graph
9
is the ideal output spectrum of a third order modulator in the case of a pulsed input, in which the modulator has ideal characteristics.
In fact, inside of such a band
12
, for example the band of FM signals, it is possible to note the presence of three minimum points
11
, which represent the three notch frequencies introduced by each of the three discrete filters included in the go path of the loop of the modulator.
The second graph
10
is, instead, the actual output spectrum from the same third order modulator in the case of a pulsed input when the phase errors and gain errors introduced by the devices or by the process spreads introduce a sort of shifting of the ideal output spectrum
9
to the actual output spectrum
10
.
It is just this shifting that produces the biggest problems of centering of the frequency.
As can be deduced from
FIG. 3
, when the integration operation is performed, an extra energy
13
is integrated, that is present over a selected threshold level
14
.
Level
14
is the thermal noise level, always present in the modulator, and the extra energy
13
is therefore that part of energy that is integrated unnecessarily, degrading the dynamic range.
A structure as heretofore described with reference to
FIG. 2
, has problems caused by the non ideality of the integrator filter
7
, which introduce phase errors and gain errors, depicted respectively by the coefficient b
1
and by the coefficient a
1
in the previous relationship (1), and gain and finite gain band product problems of the amplifier
8
.
Various solutions have been proposed to solve such problems, among which are techniques based on the master-slave concept, wherein there is a duplication of the circuitry so as to calibrate the first circuit in function of the errors of the second circuit, and compensating circuit techniques of the finite gain effects.
However no technique has been able to prevent the previous listed factors from causing a shifting of the center frequency of the modulator
1
or an integration of undesired energy. Therefore, the problem remains of centering the center band frequency in the most precise possible way, without a degrading of the dynamic range.
BRIEF SUMMARY OF THE INVENTION
According to an embodiment of the present invention, a method is provided, for calibrating a frequency of a sigma-delta modulator, said sigma-delta modulator having a go path and a feedback path, said go path including, in series, a resonator circuit and of an analog to digital conversion ADC block, said feedback path including a digital to analog conversion DAC block, said method comprising: a) applying an input pulse to said resonator circuit while the feedback path of said sigma-delta modulator is opened; b) measuring the oscillating frequency of the output signal from said resonator circuit in response to said pulse; c) comparing said oscillating frequency of said resonator circuit with a frequency known a priori; d) modifying said oscillating frequency of said resonator circuit proportionately, as a function of said comparing step (c).
According to another embodiment of the invention, a circuit for calibrating a frequency is provided, comprising a go path, including, in series, a resonator circuit and an analog to digital conversion ADC block, said resonator circuit comprising an integrator filter having, on its own feedback path, a variable gain amplifier. The frequency calibrating circuit further includes a circuit feedback path having a digital to analog conversion DAC block. Said variable gain of said amplifier is configured to be modified in a proportional way as a function of a comparison between an output signal frequency present at an output of said resonator circuit as a response to a pulsed signal present at an input thereof, and a frequency known a priori, while said circuit feedback path is opened.
Thanks to the present invention it is possible to make a measurement of a frequency within a shortened observation period, as compared to known methods and devices.


REFERENCES:
patent: 3824466 (1974-07-01), Olier et al.
patent: 3992670 (1976-11-01), Gittins et al.
patent: 5729230 (1998-03-01), Jensen et al.
patent: 6232901 (2001-05-01), Abbey
patent: 6271781 (2001-08-01), Pellon
patent: 6414615 (2002-07-01), Cheng

Affiliated with

Also associated with

Rating

Method for self-calibrating a frequency of a modulator... 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 for self-calibrating a frequency of a modulator..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for self-calibrating a frequency of a modulator... will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3305640