CN1585276A - Structure and method for carrying out A/D converter with multiple sampling rates and high accuracy - Google Patents
Structure and method for carrying out A/D converter with multiple sampling rates and high accuracy Download PDFInfo
- Publication number
- CN1585276A CN1585276A CN 03154275 CN03154275A CN1585276A CN 1585276 A CN1585276 A CN 1585276A CN 03154275 CN03154275 CN 03154275 CN 03154275 A CN03154275 A CN 03154275A CN 1585276 A CN1585276 A CN 1585276A
- Authority
- CN
- China
- Prior art keywords
- filter
- frequency
- analog
- adc
- framework
- 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.)
- Pending
Links
Images
Abstract
The structure is composed of the decimation filter and the FIR filter. In the decimation filter, the input high-speed data stream is processed with frequency degression procedure and is converted into low speed data stream, after being filtered off the input signal, it is rectified into the quantified noise of high frequency. The FIR filter processes the result of the decimation filter with frequency compensation of passing band to reach the better accuracy. On the other hand, the transition band is steeper, thus makes the greater attenuation of the frequency.
Description
Affiliated technical field
The present invention relates to a kind of framework and implementation method about multi-sampling rate high-precision analog digital quantizer (ADC).
Background technology
At present, in most of analog-digital converters (ADC) were used, digital circuit only adopted frequency reducing filter (Decimation filter), and its sample rate is changeless.Be not smooth in passband by frequency reducing filter (Decimation filter) transfer function simultaneously, it is very big to decay, and gain is successively decreased gradually at frequency domain, can not keep identical gain, the applied environment that is unsuitable for having relatively high expectations in the passband of needs; Finite impulse response filter (FIRflter) just in time can remedy the defective of the former transfer function on frequency spectrum, it is very little to accomplish that transfer function changes in passband, as be controlled at ± 0.1dB, this is essential in accurate measurement (Metering) and advanced audio (HIFI-audio) application facet.But finite impulse response filter (FIRfilter) complex structure can not adopt finite impulse response filter (FIR filter) that high-frequency input signal is carried out repeatedly frequency reducing separately.
Summary of the invention
The present invention is adopting frequency reducing filter (Decimation filter) to add on the basis of finite impulse response filter (FIR filter), utilize frequency reducing filter (Decimation filter) to realize the rate-compatible of operation result, thereby realize the variable sample rate of whole high-precision analog digital quantizer (ADC), realize multi-sampling rate, high accuracy, but improved the applied environment of this framework.
As Fig. 1, the framework of multi-sampling rate high-precision analog digital quantizer of the present invention (ADC) must comprise No. one low pass filter (Decimation filter) and a finite impulse response filter (FIR filter) at least, can also comprise a high pass filter (HPFfilter).
Modulator provides the high-speed data-flow of one tunnel one bit, and this data flow is that " 1 " has reflected the amplitude of the signal of sending into modulator with the ratio that is the number of " 0 " in a long period, thereby the analog-digital conversion function to analog input signal is provided.Certainly, the input signal of this framework also can be the multidigit metadata streams in other sources.Low pass filter (Decimation filter) is done down conversion process with a bit high-speed data-flow of modulator input, convert multidigit unit low rate data streams to, resist simultaneously and mix repeatedly, filter the quantizing noise of the quilt " noise shaping " of modulator input signal to high band.High pass filter (HPF filter) is done certain decay to the direct current biasing composition that analogue device in the analog modulator brings.Finite impulse response filter (FIR filter) is made low-pass filtering to the low speed multidigit unit that is low pass filtering device (Decimation filter) and converts to, analyze from frequency domain, transition band to input signal further decays on the one hand, makes it more precipitous, thereby reaches better precision; On the other hand, the decay in passband compensates at low pass filter (Decimation filter), and the total spectral response of system is remained on about 0dB at passband, and fluctuation (ripple) is less than positive and negative 0.1dB.
Input signal is through dual stage filter on time domain, the convolution that can regard the total transfer function of signal and dual stage filter on the frequency domain as, so in the input signal of this framework, bring noise into, the transfer function of capital and whole framework is convolution and the noise of the frequency range of result outside passband will be subjected to great decay mutually, sees Fig. 2.From the passband amplifier section of Fig. 2, can see, see Fig. 3, if not passing through finite impulse response filter (FIR filter) handles, the passband higher-frequency of product on frequency domain partly can have the decay that should not occur with respect to the counterpart of input signal, this will influence the range of application of analog-digital converter (ADC).
The 1-bit data flow of actual modulator output is observed from frequency spectrum, and the white noise that quantizes to bring is arrived the upper frequency section by " noise shaping ".If directly use, because the frequency spectrum convolution provides the stack of all frequencies on the frequency domain and has subtracted each other, the high-frequency noise of modulator then might be transformed into the low-frequency range of being concerned about so, thereby the noise platform of the low-frequency range of product is lifted to very high, makes result's signal to noise ratio variation.In actual framework as shown in Figure 1, after will being low pass filtering device (Decimation filter) low-pass filtering, the bit data flow that modulator produces imports finite impulse response filter (FIR filter) again.Through the multidigit metadata streams after low pass filter (Decimation filter) low-pass filtering, its high frequency quantizing noise has been attenuated to very low, but the upper frequency composition in the passband is attenuated relatively, need finite impulse response filter (FIR filter) that it is compensated, make to be attenuated part and to be promoted accordingly.
The present invention has improved the signal to noise ratio of operation result under the prerequisite that does not increase hardware, simultaneously in passband, different frequency becomes that the change in gain of sub-signal is minimum, the variable sample rate of operation result.During for multi-channel analog digital quantizer (ADC) signal processing, the method for available time-sharing multiplex is transformed this system, and circuit need not change too big.
Adopt the several most important advantage of this framework to be:
1. higher precision.Input signal carries out low-pass filtering with low pass filter (Decimation filter) earlier, in-band noise and out-of-band noise all have decay by a relatively large margin, use finite impulse response filter (FIR filter) to carry out meticulousr filtering again, make transition band more precipitous, and the frequency that is attenuated of passband is compensated.The noise platform of end product is compressed on very low value, and all frequency contents in the whole passband keep gain constant simultaneously.Can also adopt high pass filter (HPF filter) to filter direct current or extremely low frequency composition as required.
2. simple in structure.Owing to adopt earlier low pass filter (Decimation filter) that input signal is carried out by a relatively large margin frequency reducing, after the finite impulse response filter (FIR filter) that connects relatively just very little to the requirement of frequency reducing, only the frequency compensation function that low-pass filtering and passband are attenuated frequency need be finished, just this framework can be realized with less hardware; And, for multi-channel analog digital quantizer (ADC), can adopt the mode of time-sharing multiplex, hardware need not increase too much and can realize.
3. output result's rate-compatible can increase range of application.
Description of drawings
Fig. 1 is the structured flowchart of multi-sampling rate high-precision analog digital quantizer of the present invention (ADC), and HCLK is meant the sample frequency of the high-speed data-flow of modulator output; LCLK is meant the output result's of this framework target sample frequency, and LCLK is variable;
Fig. 2 is the spectrogram of low pass filter (Decimation filter);
Fig. 3 is the enlarged drawing of the spectrogram medium and low frequency part of low pass filter (Decimation filter);
Fig. 4 is the schematic diagram of the embodiment of this framework in the HIFI sound equipment.
Embodiment
As shown in Figure 4, the HIFI sound equipment is made of analog-digital converter (ADC), digital analog converter (DAC) and middle three unit of digital signal processor (DSP).Analog-digital converter (ADC) by this framework of 2 road time-sharing multiplexs, changes into 2 road audio analog signals first digital signal of high-precision multidigit, and guarantees that the noise of introducing is enough little; Digital signal processor (DSP) is done the audio processing with the result of analog-digital converter (ADC), makes it to have different acoustics; At last, the digital signal after digital analog converter (DAC) is handled digital signal processor (DSP) changes into audio analog signals again.
The invention provides a kind of framework and implementation method about multi-sampling rate high-precision analog digital quantizer (ADC), especially a kind of with cumulative error filter-divider (SDM, being ∑-△ modulator) high-speed data-flow that produces changes into low speed multidigit metadata, realizes the multi-sampling rate and the high accuracy of analog-digital converter (ADC).
Claims (3)
1, a kind of framework about multi-sampling rate high-precision analog digital quantizer (ADC) is characterized in that: this analog-digital converter (ADC) must comprise No. one low pass filter (Decimation filter) and a finite impulse response filter (FIR filter) at least.
2, the framework of analog-digital converter according to claim 1 (ADC) is characterized in that: this analog-digital converter (ADC) can also comprise 1 high pass filter (HPFfilter).
3, the implementation method of the framework of multi-sampling rate high-precision analog digital quantizer according to claim 1 and 2 (ADC), it is characterized in that: the bit high-speed data-flow that low pass filter (Decimation filter) will be imported is done down conversion process, convert multidigit unit low rate data streams to, resist simultaneously and mix repeatedly, filtered the quantizing noise of the quilt " noise shaping " of modulator input signal to high band, high pass filter (HPF filter) is done certain decay to the direct current biasing composition that analogue device in the analog modulator brings, finite response coefficient filter (FIR filter) is done the band connection frequency compensation to the result of low pass filter (Decimation filter), output to low pass filter (Decimation filter) compensates at the higher-frequency composition of passband, reach better precision, on the other hand, make result's excessive band more precipitous, make the outer frequency content of band of product that very big decay be arranged again.The convolution of regarding signal on the frequency domain multiplying each other of signal on the time domain as, so when total transfer function of two low pass filters of this framework low frequency (being passband) scope in if there is decay to exist, thereby it all can with input signal convolution and in its product, make useful low-frequency component be subjected to certain decay to make the input signal variation mutually.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 03154275 CN1585276A (en) | 2003-08-21 | 2003-08-21 | Structure and method for carrying out A/D converter with multiple sampling rates and high accuracy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 03154275 CN1585276A (en) | 2003-08-21 | 2003-08-21 | Structure and method for carrying out A/D converter with multiple sampling rates and high accuracy |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1585276A true CN1585276A (en) | 2005-02-23 |
Family
ID=34597959
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 03154275 Pending CN1585276A (en) | 2003-08-21 | 2003-08-21 | Structure and method for carrying out A/D converter with multiple sampling rates and high accuracy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1585276A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101197800B (en) * | 2006-12-07 | 2012-02-08 | 索尼株式会社 | Information processing device and information processing method |
| CN101207372B (en) * | 2007-12-04 | 2012-06-06 | 中兴通讯股份有限公司 | Apparatus and method for implementation of fixed decimal sampling frequency conversion |
| CN101563847B (en) * | 2006-12-19 | 2012-09-19 | 艾利森电话股份有限公司 | Fast, high resolution digital-to-analog converter |
| CN101329870B (en) * | 2008-08-01 | 2012-12-12 | 威盛电子股份有限公司 | Audio encoder and related electronic device |
| CN109428601A (en) * | 2017-08-25 | 2019-03-05 | 联发科技股份有限公司 | Improve the system of signal processing |
-
2003
- 2003-08-21 CN CN 03154275 patent/CN1585276A/en active Pending
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101197800B (en) * | 2006-12-07 | 2012-02-08 | 索尼株式会社 | Information processing device and information processing method |
| CN101563847B (en) * | 2006-12-19 | 2012-09-19 | 艾利森电话股份有限公司 | Fast, high resolution digital-to-analog converter |
| CN101207372B (en) * | 2007-12-04 | 2012-06-06 | 中兴通讯股份有限公司 | Apparatus and method for implementation of fixed decimal sampling frequency conversion |
| CN101329870B (en) * | 2008-08-01 | 2012-12-12 | 威盛电子股份有限公司 | Audio encoder and related electronic device |
| CN109428601A (en) * | 2017-08-25 | 2019-03-05 | 联发科技股份有限公司 | Improve the system of signal processing |
| US11626858B2 (en) | 2017-08-25 | 2023-04-11 | Mediatek Inc. | System improving signal handling |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102714503B (en) | Converter system, continuous-time oversampled converter and conversion method | |
| Naus et al. | A CMOS stereo 16-bit D/A converter for digital audio | |
| Lee | for High Resolution Oversampling A/D Converters | |
| Petraglia et al. | High-speed A/D conversion incorporating a QMF bank | |
| US4862169A (en) | Oversampled A/D converter using filtered, cascaded noise shaping modulators | |
| US5124705A (en) | Analog-to-digital signal converter comprising a multiple sigma-delta modulator | |
| US7453382B2 (en) | Method and apparatus for A/D conversion | |
| CN1585276A (en) | Structure and method for carrying out A/D converter with multiple sampling rates and high accuracy | |
| CN111900992A (en) | Analog-digital converter based on incremental modulation | |
| EP0229764B1 (en) | Interpolative analogue to digital converter | |
| CN1110901C (en) | Analog-to-digital conversion apparatus | |
| Stewart | An overview of sigma delta ADCs and DAC devices | |
| JP3115264B2 (en) | Digital / analog converter | |
| CN116913294A (en) | PCM signal processing method and device, digital audio system and electronic equipment | |
| CN1112777C (en) | Signal processing method and device | |
| CN108683420A (en) | A kind of lossless DSD signal number raising frequency algorithms | |
| WO2005036757A1 (en) | Delta sigma modulator with integral decimation | |
| CN1300466A (en) | Converter | |
| JPH11274937A (en) | Delta-sigma modulator an band-pass filter adopting the delta-sigma modulator | |
| CN117526957B (en) | Analog-to-digital converter with optimal quantization bit number | |
| CN117459067A (en) | Delta-Sigma digital-analog converter | |
| CN1617455A (en) | Method and circuit for effective conversion of PCM signals into PWM signals | |
| CN2742680Y (en) | Quantized noise shaping modulator | |
| US20060049970A1 (en) | Sigma-delta modulation | |
| Hummels et al. | A compensation technique for sigma-delta analog-to-digital converters |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |