JPS6150423A - Digital-analog converter - Google Patents

Abstract

PURPOSE:To improve the accuracy of D/A conversion by applying over-sampling to a digital signal to be subject to D/A conversion at a frequency nf so as to improve the frequency-phase characteritic at a signal having high frequency and subtract sufficiently dither up to a high frequency. CONSTITUTION:A digital signal of sampling frequency (f) and bit number N is inputted from a terminal 1 and a sampling pulse in a frequency nf being n-time of the frequency (f) is inputted from a terminal 4. An over-sampling circuit 2 applies over-sampling in frequency nf to the input signal. A dither from a dither generator 5 is sampled in the frequency nf similarly by an A/D converting circuit 6 to form a digital signal and the both are superimposed by an adder 3. An output signal of the adder 3 is converted into an analog signal by the 1st D/A conversion circuit 7 having (N+K)-bit more than the N-bit in bit number. A signal obtained by converting into an analog signal at the 2nd D/A converter circuit 8 from the digitize dither is subtracted from the signal converted into the analog signal by a subtractor 9 and the analog signal subtracted with the dither is obtained at a terminal 10.

Description

DETAILED DESCRIPTION OF THE INVENTION Field of Industrial Application The present invention relates to a method for processing digital signals, and improves the accuracy of a digital-to-analog conversion (D/A conversion) device that converts a digital signal into an analog signal. The present invention relates to an improvement of a D/A conversion, calculation, and subtraction method in which white noise called dither is superimposed on a digital signal.

Conventional technology The technology to improve the accuracy of D/A conversion in digital signal processing has been a major issue than before, and various circuit improvements have been attempted, but none have achieved sufficient accuracy. I couldn't.

For example, in order to convert so-called quantization noise in digital signal processing into white noise, a processing method is known in which white noise called dither is superimposed on the digital signal and the dither is subtracted after D/A conversion. The method is that if the sampling frequency is not high enough, the phase characteristics at high frequencies will deteriorate, making it impossible to perform complete subtraction at high frequencies during dither subtraction, resulting in high precision./A
The major drawback is that it cannot be converted.

Embodiment The present invention eliminates the above-mentioned drawbacks in D/A conversion.
A conversion v2@ will be described. In Figure 4, a digital signal with sampling frequency f and number of bits N is input from (1). From (4), the frequency nf is n times the frequency f.
The sampling pulse is manually generated. An oversampling circuit (2) oversamples the digital signal at a frequency nf. The dither from the dither generator (5) is sampled at the same frequency nf by the D/A conversion circuit (6) to form a digital signal, and this digitized dither and the oversampled signal are superimposed by the adder (3). let The output signal of the adder (3) is converted into an analog signal by a first D/A conversion circuit (7) of N+K bits, which has more bits than N bits. Here, the reason why the D/A conversion circuit with more bits than N bits converts it into an analog signal is because the oversampled signal and the digitized dither are superimposed in the adder (3), but normally the number of bits is more than N bits. 1 bit more N+1 bit D/
An A conversion circuit is sufficient. The first D/A conversion circuit (
When the subtracter (9) subtracts the signal converted to analog by the digitized dithered second D/A conversion circuit (8) from the signal converted to analog signal in step 7), (lO) is obtained.
An analog signal with dither subtracted is obtained.

Effects of the Invention As is clear from the above, since the digital signal to be D/A converted is oversampled at the frequency nf, the frequency phase characteristics at high frequencies of the signal are good, and redither subtraction is sufficiently performed up to high frequencies. . Therefore, it is possible to significantly improve the accuracy of D/A conversion, and it is possible to provide a high-performance/A/A conversion system.

[Brief explanation of drawings]

FIG. 1 shows a block diagram of a digital-to-analog converter according to the present invention. 2... Oversampling circuit, 3...
Adder, 5... Dither generator, 6... Analog-digital conversion circuit, 7... First digital-analog conversion circuit, 8... Second digital -Analog conversion circuit, 9...subtractor. Procedural amendment (March 14, 1920)

Claims (1)

[Claims] An oversampling circuit that oversamples a digital signal with a sampling frequency f and a number of bits N at a frequency nf that is n times higher; a dither generator; and a dither generator that samples output dither from the dither generator at a frequency nf. an analog-to-digital conversion circuit; an adder for adding the oversampled digital signal and the dither sampled at a frequency nf; a first digital-to-analog conversion circuit with N+K bits, which is one bit more; a second digital-to-analog conversion circuit for converting the dither sampled at a frequency nf into an analog signal; and a first digital-to-analog conversion circuit. A digital-to-analog conversion device comprising: a subtracter that subtracts the output signal of the second digital-to-analog conversion circuit from the output signal.