CN214591387U - Novel audio frequency DAC noise shaping controller - Google Patents

Novel audio frequency DAC noise shaping controller Download PDF

Info

Publication number
CN214591387U
CN214591387U CN202120694468.0U CN202120694468U CN214591387U CN 214591387 U CN214591387 U CN 214591387U CN 202120694468 U CN202120694468 U CN 202120694468U CN 214591387 U CN214591387 U CN 214591387U
Authority
CN
China
Prior art keywords
data
zero
noise shaping
delay
output
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202120694468.0U
Other languages
Chinese (zh)
Inventor
李军
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Individual
Original Assignee
Individual
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Individual filed Critical Individual
Priority to CN202120694468.0U priority Critical patent/CN214591387U/en
Application granted granted Critical
Publication of CN214591387U publication Critical patent/CN214591387U/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Compression, Expansion, Code Conversion, And Decoders (AREA)

Abstract

The utility model relates to a noise shaping field indicates a novel audio frequency DAC noise shaping controller especially. The greatest benefit of using an overflow quantizer in this application is that the shaper does not need to take into account the problem of system stability, which is not achieved in conventional noise shaping, which must prevent signal oscillation by controlling the output amplitude or phase of the integrator. In this design, the order of noise shaping is reflected in the bandwidth of the output data of the overflow quantizer. The one-to-one correspondence between the shaping order and the output data bandwidth of the shaper is an inherent attribute of the overflow type quantizer, so that a satisfactory result can be easily obtained by only selecting a proper order and well controlling the final data bandwidth in the design. The new technology can make the noise shaping framework simpler, the efficiency higher, the stability better, and the fidelity of the sound higher in practice.

Description

Novel audio frequency DAC noise shaping controller
Technical Field
The utility model relates to a noise shaping field indicates a novel audio frequency DAC noise shaping controller especially.
Background
The noise shaping technology is the mainstream technology of the current audio DAC, the noise shaping technology dominates almost the whole audio market with low cost and simple architecture, and the noise shaping technology is widely applied to other fields. But ideal noise shapers are unstable at or above third order and can oscillate.
Therefore, a DAC noise shaping controller is proposed in the prior art, which does not need to consider the system stability at three or more levels, and can avoid the generation of oscillation,
disclosure of Invention
In order to solve the problems, the utility model provides a novel audio DAC noise shaping controller, which solves the stability problem well without sacrificing the performance; meanwhile, the noise shaping framework is simpler, the efficiency is higher, the stability is better, and the fidelity of sound is higher in practice.
In order to achieve the above object, the utility model adopts the following technical scheme: a novel audio DAC noise shaping controller comprises a noise preprocessing module, a noise shaping module and a noise shaping module, wherein the noise preprocessing module is used for preprocessing original input signal data; a zero-delay integrator which adds the current data and the data of the previous clock cycle and outputs the data of the current clock cycle; the overflow type quantizer outputs data exceeding the part with the limited bit width when the input data exceeds the limited bit width set in the overflow type quantizer; the delayer delays the input data by one clock period and then realizes output; the accumulator is provided with two data receiving ends, and the signal data received by the two data receiving ends are added and output; the method comprises the steps that original input signal data are split into high-band bandwidth data and low-band bandwidth data through a noise preprocessing module; the high-section bandwidth data is processed by a delayer and then output to a data receiving end of the accumulator, the low-section bandwidth data is sequentially subjected to integration processing by a plurality of zero delay integrators, passes through another delayer and is output to an overflow type quantizer for processing, and then the overflow type quantizer outputs the data to another data receiving end of the accumulator.
Furthermore, the output data of the overflow type quantizer is fed back to each zero delay integrator, and each signal data input to the zero delay integrator is subtracted from the output data fed back by the overflow type quantizer, and then enters zero delay integration for integration processing.
Furthermore, the noise preprocessing module is an amplitude limiting module, and the amplitude limiting module multiplies the original input signal data by a built-in set coefficient and reduces the amplitude of the original input signal data according to the coefficient proportion.
Furthermore, a plurality of zero-delay integrators are connected in series, low-section bandwidth data is connected with the zero-delay integrator positioned at the head end, and the zero-delay integrator positioned at the tail end is connected with the overflow type quantizer through the delay unit.
Further, the number of the zero-delay integrators is n, and the number of the zero-delay integrators corresponds to the bandwidth of the output data of the overflow quantizer.
The beneficial effects of the utility model reside in that:
1. the greatest benefit of using an overflow quantizer in this application is that the shaper does not need to take into account the problem of system stability, which is not achieved in conventional noise shaping, which must prevent signal oscillation by controlling the output amplitude or phase of the integrator. In this design, the order of noise shaping is reflected in the bandwidth of the output data of the overflow quantizer. The one-to-one correspondence between the shaping order and the output data bandwidth of the shaper is an inherent attribute of the overflow type quantizer, so that a satisfactory result can be easily obtained by only selecting a proper order and well controlling the final data bandwidth in the design.
2. The new technology can make the noise shaping framework simpler, the efficiency higher, the stability better, and the fidelity of the sound higher in practice.
Drawings
Fig. 1 is a schematic block diagram of an audio DAC noise shaping controller according to the present application.
Fig. 2 is a detailed functional block diagram of an audio DAC noise shaping controller in a particular embodiment.
The reference numbers illustrate: the device comprises a zero delay integrator 1, a delayer 2, an overflow type quantizer 3, an accumulator 4 and an amplitude limiting module 5.
Detailed Description
Referring to fig. 1, the present invention relates to a novel audio DAC noise shaping controller, which includes a noise preprocessing module for preprocessing the original input signal data; a zero-delay integrator 1, wherein the zero-delay integrator 1 adds the current data and the data of the previous clock cycle and outputs the data of the current clock cycle; (wherein, in the present application, mainly the order 3 or more than 3 is referred to, so in the present application, 3 or more than 3 zero delay integrators 1 are used), when the input data exceeds the defined bit width set in the overflow quantizer 3, the overflow quantizer 3 outputs the data exceeding the defined bit width part; the delayer 2 delays the input data by one clock period and then realizes output; an accumulator 4, wherein the accumulator 4 has two data receiving ends, and the signal data received by the two data receiving ends are added and output; the method comprises the steps that original input signal data are split into high-band bandwidth data and low-band bandwidth data through a noise preprocessing module; the high-band bandwidth data is processed by the delayer 2 and then output to a data receiving end of the accumulator 4, the low-band bandwidth data is sequentially subjected to integration processing by the plurality of zero delay integrators 1 and is output to the overflow type quantizer 3 for processing by the other delayer 2, then the overflow type quantizer 3 outputs the data to the other data receiving end of the accumulator 4, and the accumulator 4 outputs the final data.
Further, the output data of the overflow quantizer 3 is fed back to each zero-delay integrator 1, and each signal data input to the zero-delay integrator 1 is subtracted from the output data fed back by the overflow quantizer 3, and then enters zero-delay integration for integration processing.
Further, the noise preprocessing module is an amplitude limiting module 5, and the amplitude limiting module 5 multiplies the original input signal data by a built-in set coefficient, and reduces the amplitude of the original input signal data according to the coefficient proportion.
Furthermore, a plurality of zero-delay integrators 1 are connected in series, low-section bandwidth data is connected with the zero-delay integrator 1 at the head end, and the zero-delay integrator 1 at the tail end is connected with the overflow type quantizer 3 through a delayer 2.
Further, the number of the zero-delay integrators is n, and the number of the zero-delay integrators corresponds to the bandwidth of the output data of the overflow quantizer.
Referring to FIG. 2, one embodiment is illustrated below:
wherein the original input signal data is 24-bit input signal samples at the beginning, and each 24-bit (a + bbit) input signal sample is first multiplied by 0.9375 ((2)a+b-2a)/2a+b) The process is to prevent the original input signal data from being too large, which causes the output data of the noise shaper to overflow and causes the output data to overflow mistakenly.
The scaled-down 24-bit data is then divided into 9-bit high data and 15-bit low data, and the 9-bit high data is delayed by the delay 2 for one clock cycle and then fed into the final summing accumulator 4. The data with 15 bits lower is processed by a 5-order noise shaping (5 zero-delay integrators 1) to generate 5-bit output data, and the 5-bit output data is sent to the final accumulator 4 to be added with the data with 9 bits higher to form final output data.
The working principle of the 5-order noise shaper for the output data of low 15bit is as follows: the 15bit input data is first subjected to 5 successive integrations and subtracted from the output data of the overflow quantizer 3 before each integration. The whole process is as follows: the 15bit input data is first subtracted from the output data of the overflow quantizer 3 to obtain the error signal of the input data and the output data, this error signal is integrated as the input signal of the integrator, the output signal of the zero-delay integrator 1 is used as the input signal of the next stage and then subtracted from the output data of the overflow quantizer 3 to obtain the error signal of the input data and the output data again, this obtained error signal is used as the input signal of the second stage integrator and integrated by the second integrator, the output of the second integrator is used as the input signal of the next stage again, and the whole integration process is repeated 5 times in cycles. The 5 th integrated data is sent to a delayer 2, the data after passing through the delayer 2 is sent to an overflow type quantizer 3 of the application, the overflow type quantizer 3 is used for reserving the data higher than 15bit as output data, and the data of 15bit and below are abandoned.
After passing through the specially designed overflow type quantizer 3, the high 5-bit data output by noise shaping has a high corresponding relation with the initial 15-bit input data, but is not completely equivalent to the input signal and has a certain error, the magnitude of the error has a strict corresponding relation with the order of the shaper, and the higher the order of noise shaping is, the smaller the error is. Therefore, the purpose of noise shaping is to obtain low-precision data after a series of numerical calculations, and then replace the original input data with the low-precision data, thereby achieving the purpose of reducing the precision of the final output data.
It is particularly emphasized that the greatest benefit of the overflow quantizer 3 is that the shaper does not need to take into account the problem of system stability. This is not achieved in conventional noise shaping, which must prevent signal oscillation by controlling the output amplitude or phase of the integrator, and even forms a subject of special study in the design and study of higher-order noise shaping. In the present design, the order of noise shaping is reflected in the bandwidth of the output data of the overflow quantizer 3. Specifically, the number n of the zero delay integrators corresponds to the bandwidth of the data output by the overflow type quantizer, the number n of the zero delay integrators forms n-order noise shaping, and the bandwidth of the data output by the overflow type quantizer 3 is nbit; for example, the bandwidth of the output data of the 5 th order noise shaper is 5bit, if we reduce the order of noise shaping to 4 th order, the bandwidth of the output data becomes 4bit, and the one-to-one relationship between the shaping order and the bandwidth of the output data of the shaper is an inherent property of the overflow quantizer 3. A satisfactory result can be easily obtained by selecting a proper order and controlling the final data bandwidth well in the design.
The noise shaping structure designed in this way is simple, and it can be seen from the above example that most of the numerical calculation is for data with a low bit of 15 bits, and a data value with a high bit of 9 bits is delayed by one clock. The last 9BIT data is directly sent to the PWM modulator, the main clock frequency of the PWM modulator with the output carrier frequency of 384K is 196.6MHz, the speed is very easy to realize by the prior art, and the DEMO of XILINX version is that 24BIT input data is integrated into 9BIT data which is modulated into a PWM signal of 384kHz by the PWM modulator.
In terms of tone quality, just like the structure, under the D class amplifier mode of 8 times of oversampling, 5-level noise shaping, 384kHz and PWM output, the high fidelity requirement in the true sense can be achieved, and real sound like the scene absolutely allows you to map deeply. Meanwhile, the carrier frequency of 384kHz is just the frequency used most by the most popular class D amplifier at present, which is just the same as the carrier frequency of the class D amplifier with high guarantee.
The above embodiments are only for describing the preferred embodiments of the present invention, and are not intended to limit the scope of the present invention, and various modifications and improvements made by the technical solution of the present invention by those skilled in the art are all within the scope of the present invention as defined by the claims.

Claims (5)

1. The utility model provides a novel audio DAC noise shaping controller which characterized in that:
the device comprises a noise preprocessing module, a noise detection module and a noise detection module, wherein the noise preprocessing module is used for preprocessing original input signal data;
a zero-delay integrator which adds the current data and the data of the previous clock cycle and outputs the data of the current clock cycle;
the overflow type quantizer outputs data exceeding the part with the limited bit width when the input data exceeds the limited bit width set in the overflow type quantizer;
the delayer delays the input data by one clock period and then realizes output;
the accumulator is provided with two data receiving ends, and the signal data received by the two data receiving ends are added and output;
the method comprises the steps that original input signal data are split into high-band bandwidth data and low-band bandwidth data through a noise preprocessing module; the high-section bandwidth data is processed by a delayer and then output to a data receiving end of the accumulator, the low-section bandwidth data is sequentially subjected to integration processing by a plurality of zero delay integrators, passes through another delayer and is output to an overflow type quantizer for processing, and then the overflow type quantizer outputs the data to another data receiving end of the accumulator.
2. The novel audio DAC noise shaping controller of claim 1, wherein: the output data of the overflow type quantizer is fed back to each zero delay integrator, and each signal data input to the zero delay integrator is subtracted from the output data fed back by the overflow type quantizer and then enters zero delay integration for integration processing.
3. The novel audio DAC noise shaping controller of claim 1, wherein: the noise pre-processing module is an amplitude limiting module which multiplies the original input signal data by a built-in set coefficient and reduces the amplitude of the original input signal data according to the coefficient proportion.
4. The novel audio DAC noise shaping controller of claim 1, wherein: the plurality of zero-delay integrators are connected in series, the low-section bandwidth data is connected with the zero-delay integrator positioned at the head end, and the zero-delay integrator positioned at the tail end is connected with the overflow type quantizer through the delay unit.
5. The novel audio DAC noise shaping controller of claim 1 or 4, wherein: the number of the zero-delay integrators is n, and the number of the zero-delay integrators corresponds to the bandwidth of the output data of the overflow quantizer.
CN202120694468.0U 2021-04-06 2021-04-06 Novel audio frequency DAC noise shaping controller Active CN214591387U (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202120694468.0U CN214591387U (en) 2021-04-06 2021-04-06 Novel audio frequency DAC noise shaping controller

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202120694468.0U CN214591387U (en) 2021-04-06 2021-04-06 Novel audio frequency DAC noise shaping controller

Publications (1)

Publication Number Publication Date
CN214591387U true CN214591387U (en) 2021-11-02

Family

ID=78323263

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202120694468.0U Active CN214591387U (en) 2021-04-06 2021-04-06 Novel audio frequency DAC noise shaping controller

Country Status (1)

Country Link
CN (1) CN214591387U (en)

Similar Documents

Publication Publication Date Title
US7190288B2 (en) Look-up table delta-sigma conversion
US20060282185A1 (en) Device and method for signal processing
US9337821B2 (en) System and method for generating a pulse-width modulated signal
US6384761B1 (en) Second and higher order dynamic element matching in multibit digital to analog and analog to digital data converters
KR20040071289A (en) Method and apparatus for generating a pulse width modulated signal
WO2004107563A2 (en) High-efficiency amplifier, converter and methods
JPH04261225A (en) Analog-to-digital converter provided with multiple sigma-delta modulator
JPH07105724B2 (en) Digital-analog conversion circuit
US7782238B2 (en) Asymmetric PWM signal generator, method thereof, and data processing apparatus including the same
CN214591387U (en) Novel audio frequency DAC noise shaping controller
JP2003124812A (en) Trellis type noise shaping modulator
CN117526957A (en) Analog-to-digital converter with optimal quantization bit number
US6842131B1 (en) Delta-sigma modulator
CN108832931B (en) Delta-Sigma modulator using externally applied perturbation signal
JP2000049613A (en) Digital switching amplifier
CN112491391B (en) Interpolation filter implementation structure of audio DAC
RU2730443C1 (en) Digital-to-analogue conversion device, method, data medium, electronic musical instrument and information processing device
US10848174B1 (en) Digital filter
CN103916143B (en) The non-contour step signal processing method of very low frequency long-wave transmitter and device
CN1588806B (en) Quantizing noise shaping modulator and quantizing noise shaping method
CN103780262A (en) Digit simulation conversion device and method for differential type interpolation pulse width modulation
CN111224665B (en) Apparatus and method for reducing dynamic device matching resources in an audio digital-to-analog converter
JP7213947B2 (en) Delta-sigma modulator and communication equipment
JP3003198B2 (en) Pulse width modulator
EP1556953A1 (en) Data converter

Legal Events

Date Code Title Description
GR01 Patent grant
GR01 Patent grant