Coded data generation or conversion – Analog to or from digital conversion – Differential encoder and/or decoder
Reexamination Certificate
2000-09-05
2002-02-26
JeanPierre, Peguy (Department: 2819)
Coded data generation or conversion
Analog to or from digital conversion
Differential encoder and/or decoder
C341S131000
Reexamination Certificate
active
06351229
ABSTRACT:
BACKGROUND OF THE INVENTION
By way of background, delta sigma modulators are known to inherently generate tones, due to the characteristics of delta-sigma modulator feedback loops. The tones of a delta-sigma modulator can be thought of as low frequency envelopes of the serial stream of digital pulses produced at the output of the delta sigma modulator. The frequency and level of such tones depend on the architecture of the delta sigma modulator and the level of the input signal applied to it. Although the tone levels can be very low, i.e., many dB below the noise level, the tones nevertheless can be very detrimental to the performance of the delta sigma modulator. For example, in an audio application, a tone produced in the signal band by a delta-sigma modulator typically in the audible frequency range of 20 Hz to 20 kHz can be detected by human ears, even if the level of the tone is many dB below the noise level. Such tones are in the baseband of the delta sigma modulator, and are likely to be detrimental to its performance in audio frequency applications even if the magnitude of the in-band tone is many dB below the noise level of the output signal produced by the delta sigma modulator. This is because many humans are extremely sensitive to in-band audio frequency tones, especially if the tones persist at particular frequencies. If the tone generated at the output of the delta sigma modulator includes energy that is clustered around certain frequencies, sensitive humans can hear the tone and be annoyed by it, even though the magnitude of the tone is well below the noise level of the output signal produced by the delta sigma modulator.
One known solution to the problem of in-band DC-dependent tones produced by delta sigma modulators is to add a DC offset at or near the input of the delta sigma modulator, wherein the DC offset magnitude is large enough to shift the in-band tone out of the baseband. One prior art reference disclosing this technique is commonly assigned U.S. Pat. No. 5,835,038 by Nakao et al. However, for this technique to be effective, an AC input signal applied to the delta sigma modulator can not have a DC component that might counteract or cancel an introduced opposite-polarity DC offset. If the input signal includes a DC component, the input signal must be AC coupled to the input terminal of the delta sigma modulator, which usually is inconvenient. For a delta sigma modulator having a differential input, there are a number of additional reasons why it would be desirable to have a better approach to breaking up tones or shifting them out of the baseband than just adding a DC offset or a DC dither. For example, an input signal of magnitude close to the DC offset but of opposite polarity could bring a tone that has been shifted out of the baseband back into the baseband. As another example, a very slowly-varying AC signal can have instantaneous values of magnitude comparable to the DC offset end of opposite polarity to the DC offset. This condition creates a short-duration tone.
Another approach to the problem of in-band DC-dependent tones produced by a delta sigma modulator is to introduce dither in order to de-correlate the modulator output, i.e., to break up the tones. The magnitude of a dither signal needs to reach a certain level in order to accomplish the objective of breaking up tones generated by the delta sigma modulator. However, when the dither signal has the needed level and at the same time the input signal applied to the delta sigma modulator has a large magnitude, the combination can cause instability in the delta sigma modulator. That is, a large amplitude dither signal may cause instability in the delta sigma modulator when an integrating loop of the delta sigma modulator is receiving a large magnitude input signal. To avoid this, it may be necessary to restrict the input signal range of the delta sigma modulator, which is undesirable because restricting the input signal range may limit the signal-to-noise ratio that can be achieved, especially for a high order delta sigma modulator wherein the higher the order of a delta sigma modulator, the smaller the usable input signal range must be to avoid instability problems. Typically, simple addition of dither can reduce the input range of a delta sigma modulator by a large amount, especially in a high order delta sigma modulator which would inherently already have a small stable input range.
In order to ensure loop stability in a delta sigma modulator while maintaining effective dither, the dither has to be applied “dynamically”. The known concept of dynamic dithering is to apply maximum dithering when the magnitudes of the input signals applied to the delta sigma modulator are small, because this is when undesirable tones appear in the baseband and there is a strong need for dithering to “break up” the tones. However, when the magnitude of the input signal gets larger, the magnitude of the dithering signal is reduced in order to prevent the combination of the large magnitude input signal and the large magnitude dynamic dither signal from causing instability in the delta sigma modulator. The closest prior art dynamic dithering techniques involve use of a multi-bit level estimator/quantizer and a multi-level dither injector. These multi-level circuits are necessary to cause the magnitude of the dither signal to change gradually.
“Prior art”
FIGS. 1 and 2
herein illustrate such known adaptive or dynamic dithering techniques. Basically, what is accomplished using these techniques is to estimate the level of the input signal applied to the delta sigma modulator, either by directly estimating the level of the input signal as shown in prior art
FIG. 2
, or by indirectly estimating the input level by directly estimating the output level produced by the delta sigma modulator as shown in prior art FIG.
1
. Prior art
FIG. 1
illustrates a conceptual design that would be difficult to implement because of various practical difficulties, especially latency in the digital filtering circuitry thereof. Use of the first-stage destination filter compounds this difficulty because by definition the destination filter is clock and updated at a divided-down frequency, extending the duration of the latency. Prior art
FIG. 2
corresponds to either FIG. 1 or FIG. 2 of prior art U.S. Pat. No. 5,745,061 by Norsworthy et al., entitled “Method of Improving the Stability of a Sigma-Delta Modulator Employing Dither”. Also see U.S. Pat. No. 5,144,308 by Norsworthy. Note that the dynamic dithering systems of prior art
FIGS. 1 and 2
control the magnitude of the dynamic dither signal.
The techniques utilized in both of the above mentioned references therefore arbitrarily the determine the boundaries of “transition points” at which the dither magnitude is changed, were and therefore cause the overall noise level at the output of the delta sigma modulator to undergo sudden increases across the arbitrary boundaries. The dynamic dithering techniques of both prior art
FIGS. 1 and 2
require costly multi-bit DACs and also require costly multi-bit quantizers or ADCs to control of the levels produced by the estimators, wherein the number of bits of the multi-bit DAC must be commensurate with a number of levels produced by the nonlinear quantizer, and additional circuitry is required to control the DAC.
Furthermore, in order for such prior dynamic dithering techniques to be effective in decorrelating the modulator output (i.e., in breaking up the tones), the amplitude of the dither signal has to be quite large. The need to use multiple-bit DACs and multiple-bit ADCs or quantizers to control the dither in the above prior art delta sigma modulators with dynamic dither circuitry may add substantially to the cost of the delta sigma modulators.
Various pseudo-random sequence generators are known to those skilled in the art. For example, the pseudo-randomization of the chopper clock signal frequency effectively “spreads” chopper clock noise energy throughout the frequency spectrum, and thereby reduces the intermodulation between the ampl
Brady W. James
Jean-Pierre Peguy
Swayze, Jr. W. Daniel
Telecky , Jr. Frederick J.
Texas Instruments Incorporated
LandOfFree
Density-modulated dynamic dithering circuits and method for... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Density-modulated dynamic dithering circuits and method for..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Density-modulated dynamic dithering circuits and method for... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2959272