SU984032A1 - Method and apparatus for analogue-digital conversion - Google Patents

Description

(54) METHOD OF ANALOG-DIGITAL TRANSFORMATION AND DEVICE FOR ITS IMPLEMENTATION

The invention relates to a pulse technique, in particular, to the technique of converting anaipog signals into a digital code, and can be used to transmit messages in communication systems. . A method of analog-to-digital conversion of the input signal is implemented, implemented in the device, based on integrating the difference between the current value of the first derivative and the memorized value at the time of the sample, which is the current value of the linear approximation error. The sampling times are determined by the value of the integral of a given approximation error. The disadvantage of this method is the greater sensitivity to low-frequency fluctuation noise associated with the necessity of the calculation, the derivative. It is known that during differentiation the interference spectrum changes, namely: the amplitudes of the spectral components will be proportional to the frequency value, i.e. white noise with a uniform distribution of the spectrum will have a triangular spectrum after differentiation. This leads to a deterioration of the signal-to-noise ratio after differentiation, since most signals have a falling spectrum. This is especially true for biomedical signals. Thus, in the signal regions with a zero (or mallet) value of the derivative, after differentiation, a derivative interference signal is formed with underlined high-frequency components. The signal-to-noise ratio in these areas becomes less than 1 and at the time of sampling a non-zero value of the derivative will be memorized (as it should be in the absence of noise and interference), which leads to a malfunction of the device. There is also known an analog-to-digital conversion method based on storing the input signal at reference points, determining the increment of the current value, and determining the quantization times to reach the threshold level. A device for implementing this method comprises a storage unit, the first input of which is connected to the input bus, and a comparator, the output of which is connected to the output bus and to the second input of the memory block. and the first input is with a threshold voltage 2 bus. The disadvantages of this device are also high sensitivity to high-frequency fluctuational noise, due to the need to calculate the derivative. The purpose of the invention is to increase noise immunity and increase the reliability of information. The goal is achieved by the method of analog-digital conversion, based on storing the input signal at the time of counting the increment of the current value of the input signal relative to its memorized value and determining the Quantization moments after reaching the threshold level, the formation of the quantization moments takes place. voltages proportional to the square of the input signal and the time over which the input signal was incremented, and their summation. Moreover, in the analog-digital conversion device containing the storage unit, the first input of the KOTODO is connected to the input bus, and the comparator, the output of which is connected to the output bus and to the second input of the storage unit, is entered in series with the bus of the threshold voltage. the first adder, the quadrant and the second adder, as well as the generator of the quadratic voltage, the input of which is connected to the output of the comparator, and the output to the second input of the second adder, the output of which is connected to the second input of the comparator, the first input of the first adder is connected to the input bus, and the second input is connected to the output of the storage unit. Figure 1 shows plots explaining the method, where the current value of the input signal is compared with the memorized value at the time of reference; Zig.2 shows the block diagram of the device that implements the proposed method. The square of the input value is summed with the quadratically varying voltage proportional to the square of the time increment. When the value of the sum of this level is reached, a sampling pulse is formed, which is used to memorize the new value of the input signal in the sampling time. The device contains a series-connected adder 1, a quad 2, a second adder 3, and a comparator 4, as well as a random-access memory 5 and a generator 6 of square-law voltage. You run the comparator 4 is connected to the control inputs of the generator 6 square-law voltage. The output of the comparator, 4, is connected to the control inputs of the generator 6 of the quadratic voltage and the RAM 5, the input of the RAM 5 is connected to the input of the adder 1, which is the input of the device, and the output - to the subtracting input of the first adder 1. The output of the generator 6 voltage is connected to the second input of the second sum of the dispenser 3. The device operates as follows. At the time of formation of the output signal by the comparator 4 (time t,), the generator of quadratic voltage 6 is reset to zero, and the current {value of the input signal (U) is rewritten into operational Ps1mb 5. The adder 1 and the square 2 carry out the formation of a voltage proportional to the square of the signal increment. which at adder 3 is summed with a voltage proportional to the square of the time increment generated by the square-voltage generator 6. This sum represents the current value of the square of the length of the segments of the linearly approximated signal, shown in Figure 1 by a solid line. When this value is reached, the value of the Upop threshold, defined in proportion to the square of unit length, triggers comparator 4, which forms the output signal (time t). Then the cycle repeats. Thus, the moment i and the sampling points are determined when the condition is fulfilled (U, .. (t.-t..j2. V ae, where the unit length of the linear approximation segment is, or (/ lU.-JU..j (i. -t.2, g, con5t, i.e., uniformly quantized over the lengths of sections of a linearly approximated signal.In sections of the input signal with a large derivative value, the sampling moments will be set to a snatch than in sections with a smaller derivative value. sampling rate and rate of change of the input signal. The proposed method in comparison with the known methods consists in reducing the redundancy of information in the original message for signals varying at a finite rate, which, in turn, improves the efficiency of using the communication channel when transmitting messages, as well as increasing the interference SUSTAINABILITY, in particular , to high-frequency fluctuating noise due to the exclusion of the differentiation operation.

In addition, the proposed method has a number of advantages enabling it to be effectively used in the transmission of information. First, the presence of a one-to-one connection between x and l, between quantization points (flX), allows only one coordinate to be transmitted. The same circumstance makes it possible to immediately determine the slope of the current signal section during linear interpolation when the received message is restored. Secondly, the proposed method makes it possible to easily reproduce a discretely transmitted signal in real time. The fact is that the known methods, having a predetermined interval between the moments of discretization, require a preliminary recording of information on an intermediate carrier (or in the operative memory,). In addition, provided there is no interference, these methods do not give sampling points in the signal sections with a constant value and, therefore, the ordinates between the connected sampling points can differ significantly from each other, which leads to the impossibility of using a stepped interpol during playback of the received message. The proposed method has the maximum diQChd1 values equal to ВВ, which, taking into account the functional connection between them, allows real-time (without delay) to restore the received discrete message. The simplest addition, which consists in comparing adjacent samples and discarding subsequent ones with equal slope (comparison by At or by), allows further compression of information for its storage and obtaining information compression characteristic of known methods (Retaining high noise immunity inherent to the proposed method In addition, the proposed method allows approximating the original analog signal piecewise-linear with equal lengths of sections. This can be used for p shenii issues the discrimination signal form or signal to analytically describe the purpose posleduyupey performing other processing.

the advantages of the proposed method should also include the simplicity of its hardware implementation. The formula of the invention 1. An anodic-digital conversion method based on storing the input signal at the time of reference, determining the increment of the current input signal relative to its memorized value and determining the quantization moments upon reaching the threshold level, characterized in that, in order to increase the noise immunity and increase the reliability information, before determining the moments of 1 quantization, 1, the formation of stresses is carried out, proportional to x squared input signal increment and time, The second was the increment of the input signal, and their summation. 2. An analog-to-digital conversion device containing a storage unit, the first input of which is connected to the input bus, and a comparator, the output of which is connected to the output bus, and to the second input of the storage unit, and the first input - with a threshold voltage bus, characterized in that it is connected in series with the first adder, quad and second adder, as well as a quadratic voltage generator, the input of which is connected to the comparator output, and the output to the second input of the second adder, whose output is connected to the second th input of the comparator, the first input of the first adder is connected to an input bus, and the second input - to the output zapomingioschego block. Sources of information taken into account in the examination 1. The author's certificate of the USSR W 574852, cl. H 03 K 13/02, 1975. 2. The author's certificate of the USSR 588625, cl. H 03 K 13/02, 1976 (prototype).

and.

Fig 2

Claims (2)

1. The method of analog-to-digital conversion, based on storing the input signal at the time of reference, determining the increment of the current input signal relative to its stored value and determining the quantization moments upon reaching the threshold level, characterized in that, in order to increase noise immunity and increase the reliability of information, before by determining the moments of quantization I, the formation of voltages proportional to the square is performed, the increment of the input signal and the time during which the increment ie the input signal, and their summation. 2. An analog-to-digital conversion device containing a storage unit, the first input of which is connected to the input bus, and a comparator, the output of which is connected to the output bus and the second input of the storage unit, and the first input - with a threshold voltage bus, characterized in that a first adder, a quadrator and a second adder are connected in series, as well as a quadratic voltage generator, the input of which is connected to the output of the comparator, and the output - with the second input of the second adder, the output of which is connected to the second m input of the comparator, and the first input of the first adder is connected to the input bus, and the second · input - with the output of the storage unit.