# US 2014/0192933 A1 · Demodulator for frequency-shift keyed signals by using the goertzel algorithm

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2014/0192933 A1
- Document ID: US20140192933A1
- Status: Published application
- Kind: A1
- Application number: 14/152,034
- Filed: 2014-01-10
- Published: 2014-07-10
- Earliest claimed: 2013-01-10
- Assignees: ABB Technology AG
- Applicants: ABB Technology AG
- Inventors: Stefan POOK
- Web page: https://patentgrep.com/patent/US20140192933A1

## CPC

- H04L 27/148: using filters, including PLL-type filters
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

Exemplary embodiments relates to a demodulator for frequency-shift keyed signals by using the Goertzel algorithm according to the general recursive calculating rule, where c re = 2 * cos  ( 2 * π * f g f s ) with the sampling frequency fs and the specified frequency fg. To reduce the mathematical complexity in feed-power-limited devices, it is proposed that the sampling frequency fs is an integral multiple of the frequency fg of a specified discrete spectral component.

## Claims

### Claim 1

A demodulator for frequency-shift keyed signals, comprising:
a processor configured to filter input signals and process the filtered signals using a Goertzel algorithm according to a general recursive calculating rule M=s[n]2+s[n−1]2+s[n]*s[n−1]*cre
where s[n]=x[n]+cre*s[n−1]−s[n−2],
M is the magnitude of a specified discrete spectral component,
x[n] are the samples of the input signal,
s[n] are state signals, and
cre is the feedback coefficient of the state signal s[n−1],
where
c
re
=
2
*
cos

(
2
*
π
*
f
g
f
s
)
with a sampling frequency fs and a specified frequency fg, and
wherein the sampling frequency fs is an integral multiple of the frequency fg of the specified discrete spectral component.

### Claim 2 (depends on 1)

The demodulator as claimed in claim 1, wherein a ratio of the sampling frequency fs to the frequency fg of the specified discrete spectral component is 3.

### Claim 3 (depends on 1)

The demodulator as claimed in claim 1, wherein a ratio of the sampling frequency fs to the frequency fg of the specified discrete spectral component is 4.

### Claim 4 (depends on 1)

The demodulator as claimed in claim 1, wherein a ratio of the sampling frequency fs to the frequency fg of the specified discrete spectral component is 6.

### Claim 5

A method for demodulating frequency-shift keyed signals, comprising:
filtering input signals in a processor;
processing, in a processor, the filtered signals using a Goertzel algorithm according to a general recursive calculating rule M=s[n]2+s[n−1]2+s[n]*s[n−1]*cre
where s[n]=x[n]+cre*s[n−1]−s[n−2],
M is a magnitude of a specified discrete spectral component,
x[n] are samples of the input signal,
s[n] are state signals, and
cre is a feedback coefficient of the state signal s[n−1],
where
c
re
=
2
*
cos

(
2
*
π
*
f
g
f
s
)
with a sampling frequency fs and a specified frequency fg, and
wherein the sampling frequency fs is an integral multiple of the frequency fg of the specified discrete spectral component.

### Claim 6 (depends on 5)

The method as claimed in claim 5, wherein a ratio of the sampling frequency fs to the frequency fg of the specified discrete spectral component is 3.

### Claim 7 (depends on 5)

The method as claimed in claim 5, wherein a ratio of the sampling frequency fs to the frequency fg of the specified discrete spectral component is 4.

### Claim 8 (depends on 5)

The method as claimed in claim 5, wherein a ratio of the sampling frequency fs to the frequency fg of the specified discrete spectral component is 6.

### Claim 9

A demodulator for frequency-shift keyed signals, comprising:
a processor configured to filter a line signal, sample the filtered signal and process samples output from the filter with a Goertzel algorithm, comparing a result of the algorithm to a threshold to determine whether the filter output is to be fedback for further processing with the Goertzel algorithm.

### Claim 10 (depends on 9)

The demodulator of claim 9, wherein the line signal is filtered at a sampling frequency having a magnitude that is six times a frequency of a specified symbol frequency of the line signal.

### Claim 11 (depends on 9)

The demodulator of claim 9, wherein the processor is configured to execute the Goertzel algorithm according to a general recursive calculating rule M=s[n]2 +s[n-1]2 +s[n] * s[n-1] * cre
where s[n] =x[n] +cre* s[n-1] - s[n-2],
M is a magnitude of a specified discrete spectral component,
x[n] are samples of the input signal,
s[n] are state signals, and
cre is a feedback coefficient of the state signal s[n−1],
where
c
re
=
2
*
cos

(
2
*
π
*
f
g
f
s
)
with a sampling frequency fs and a specified frequency fg, and
wherein the sampling frequency fs is an integral multiple of the frequency fg of the specified discrete spectral component

### Claim 12 (depends on 9)

The demodulator as claimed in claim 9, wherein a ratio of the sampling frequency fs to the frequency fg of the specified discrete spectral component is 3.

### Claim 13 (depends on 9)

The demodulator as claimed in claim 9, wherein a ratio of the sampling frequency fs to the frequency fg of the specified discrete spectral component is 4.

### Claim 14 (depends on 9)

The demodulator as claimed in claim 9, wherein a ratio of the sampling frequency fs to the frequency fg of the specified discrete spectral component is 6.