US 6,642,874 B1 · Sigma-delta data conversion employing quantization error feedback

Abstract

A sigma-delta circuit converts an analog or digital input sequence xn representing a quantity with at least p-bit resolution into an m-bit output sequence yn, where m<p<0. The sigma-delta circuit offsets the xn sequence by a first feedback sequence cn that is proportional to output sequence yn to produce a sequence an. The signal-delta circuit filters the an sequence using a filter having a transfer function H(z−1) to produce a sequence dn that is offset by a second feedback sequence rn to produce an analog or digital sequence bn. The bn sequence is then digitized or quantized to produce the lower resolution m-bit output yn sequence. To increase the signal-to-noise ratio within a signal band of interest, the sigma-delta circuit includes a quantization error feedback circuit for processing the bn and yn sequences to generate a quantization error sequence en representing a difference between values the bn and yn sequences represent, and for filtering the en sequence with a transfer function R(z−1) to produce the rn sequence. The quantization noise in the signal band can be shaped by adjusting R(z−1) without affecting the sigma-delta circuit's response to the x(n) sequence, which is controlled by H(z−1) and not influenced by R(z−1).

Claims30

Dependent claims

A method for converting a sequence of elements xn into a sequence of elements yn, wherein each element xn represents a magnitude with at least p-bit resolution and each element yn is an m-bit digital data word, wherein m and p are integers and wherein p>m>0, the method comprising the steps of:

a. providing a sequence of elements cn in response to the sequence of elements yn wherein each element cn corresponds to a separate element yn and represents similar a magnitude;
b. generating a sequence of elements an, each representing a magnitude proportional to a difference in magnitudes represented by a separate concurrently occurring pair of elements xn and cn;
c. filtering the sequence of elements an to produce a sequence of elements dn;
d. generating a sequence of elements bn in response to the sequence of elements dn and a sequence of elements rn, wherein each element bn represents with at least p-bit resolution a difference in magnitudes represented by a separate concurrently occurring pair of elements dn and rn;
e. generating each element yn in response to a corresponding one of elements bn, wherein each element yn represents with m-bit resolution a same magnitude its corresponding element of the bn sequence represents with at least p-bit resolution;
and
f. generating the sequence of elements rn as a function of a quantization error differences between magnitudes represented by corresponding pairs of elements bn and yn.

4 more elements

An apparatus for processing a sequence of elements xn, each representing a magnitude, the apparatus comprising:

first means for providing a sequence of elements cn in response to a sequence of elements yn, wherein elements yn are m-bit data words representing magnitudes with m-bit resolution, wherein concurrently occurring elements cn and yn represent similar magnitudes;
second means for generating a sequence of elements an in response to the sequences of elements xn and cn, wherein each element an represents a magnitude proportional to a difference between magnitudes represented by a separate concurrently occurring pair of elements xn and cn;
third means for filtering the sequence of elements an to produce a sequence of elements dn;
fourth means for generating a sequence of elements bn in response to the sequence of elements dn and a sequence of elements rn, wherein each element bn represents with at least p-bit resolution a magnitude proportional to a difference between magnitudes represented by a separate concurrently occurring pair of elements dn and rn, wherein m and p are integers and, wherein p>m>0,
fifth means for generating the sequence of elements yn in response to the sequence of elements bn, wherein each element yn corresponds to a separate one of elements bn and represents with m-bit resolution a same magnitude its corresponding element bn represents with at least p-bit resolution;
and
sixth means for generating the sequence of elements rn wherein each element rn sequence represents a magnitude proportional to a quantization error difference between magnitudes represented by a separate corresponding pair of elements bn and yn.

4 more elements

Dependencies are inferred from the claim text.

Sigma-delta data conversion employing quantization error feedback

Life of this patent

Sigma-delta data conversion employing quantization error feedback

Life of this patent

Abstract

Claims30