RU2099866C1 - Method for representation of analog signal - Google Patents

Abstract

FIELD: information encoding and transmission, in particular, electronic information devices, recording, broadcasting, TV equipment. SUBSTANCE: method involves generation of modulation signal as sequence of non-rectangular periodic pulses, multiplication of analog signal by modulation signal in order to achieve modulated signal, comparison of modulated signal to reference constant, generation of encoded pulses which leading edges start in moments when modulated signal and reference constant are equal, sending of encoded pulses to decoder which outputs demodulation signal which is inverse to modulation one, detection of values of demodulation signal for leading edges of encoded pulses and generation of discrete values of converted analog signal. EFFECT: increased functional capabilities. 7 dwg

Description

 The invention relates to the field of coding and transmission of information and can be used in information electronic devices, sound recording, broadcasting, television.

 In its technical essence, the closest to the claimed object is a known method (prototype) of analog-to-digital and digital-to-analog conversions [1] by which the analog signal is divided by triggering pulses into periodic time (clock) intervals and in each interval the analog signal is converted to encoded pulses in the form of a digital code. The latter is a sequence of counting pulses arranged in time in a certain way. According to another close prototype method [2, p. 232] [3, p. 125], this sequence is equidistantly spaced pulses, the number of which is proportional to the value of the encoded analog signal. The great advantage of the prototype is its noise immunity when transmitting digital information from a transmitter over communication lines or via ether to a receiver. The disadvantage of the prototype is the inevitable error of digitizing the analog signal due to the quantum nature of this operation, and there is uncertainty both in amplitude and in time. In addition, since the counter pulse generator must operate at a frequency significantly higher than the clock, there are problems converting high-frequency analog signals. In addition, digitizing and decoding devices are quite complex, which is especially manifested in an information system that includes many receivers (digital sound recording, broadcasting, television, etc.).

 Another well-known technical solution is the method of a frequency-modulated converter [2, p. 242] in which the pulse frequency is varied, the change in frequency being proportional to the change in the analog signal. In this method, the frequency is continuously associated with the amplitude of the analog signal. However, in the inverse transformation of the frequency-modulated signal, it is divided into periodic time intervals and within each interval the frequency is determined by counting pulses. In this case, firstly, an average frequency is obtained over the interval, and secondly, the number of pulses is an integer value. Thus, this type of conversion is also characterized by the uncertainty (inaccuracy) of obtaining the discrete value of the converted analog signal.

 In all methods of converting an analog signal, first, discrete values of the converted signal are obtained, tied to their time coordinates. Then, interpolation methods are applied, with the help of which discrete values are converted into a continuous signal.

 In the inventive method of converting an analog signal in the conversion range with a large minimum of zero by using synchronizing pulses, the time scale is divided into periodic intervals and in each of them a modulating signal is generated in the form of non-rectangular pulses, the minimum of which is less than or equal to more than zero, the working minimum, and the maximum is greater or equal to the working maximum, with the ratio of the working maximum to the working minimum greater than the ratio of the maximum to the minimum for the analog signal. The analog signal is multiplied by a modulating signal and a modulated signal is obtained. The modulated signal is compared with a reference constant, the value of which is less than the product of the minimum of the analog signal and the working maximum and more than the product of the maximum of the analog signal and the working minimum. Form coded pulses, the edges of which coincide with the moments of equality of the modulated signal and the reference constant. Carry out the inverse transformation, for which they generate a demodulating signal inversely proportional to the modulating signal. The demodulating signal is synchronized with the timeline of the encoded pulses by using synchronizing pulses. The values of the demodulating signal are recorded at the edges of the encoded pulses coinciding with the moments of equality of the modulated signal and the reference constant, and discrete values of the converted analog signal are obtained.

The mathematical justification of the proposed method is as follows. Let A (t) be an input analog signal, M (t) be a modulating signal, and E be a reference constant. We multiply the analog signal by a modulating one and get a modulated signal A (t) • M (t). We direct it to a solving device, with which we solve the equation
A (t) • M (t) = E (1)
We find the set of values t _r , where r 1, 2, 3. are the roots of equation (1). For each value of t _r we have the equality:
A (t _r ) • M (t _r ) = E, (2)
from which it follows:
A (t _r ) = E • M ^-1 (t _r ) ≈ C • M ^-1 (t _r ), (3)
where C is a greater than zero coefficient.

 Expressions (1) (3) show that if a modulating signal and used constants are generated in an encoding device, an equation including the analog and modulating signal along with the necessary constants is compiled, the equation is solved using the resolver, finding its roots, and the root information is sent to a decoding device in which to generate the necessary constants and a demodulating signal inversely proportional to the modulating one, to fix the values of the demodulating signal at the moments of the roots, having received them in phase equal to the phase of the root values of the modulating signal, then at the output of the decoding device we obtain discrete values of the converted analog signal proportional to the root values of the input analog signal. As a solver, we use a well-known comparator in which we compare an alternating signal with a reference constant E. To obtain a value of a demodulating signal in a decoder in a phase equal to the phase of the root value of the modulating signal, we generate synchronizing pulses in the encoder that are transmitted along with the root information in decoder. Since the instant method converts the instantaneous values of the analog signal to the moments of the roots of the equations to be solved, the root information transmitted from the encoding device to the decoding one should be pulses called coded by analogy with the prototype, having step edges at the root moments (in order to separate these fronts are from stepped fronts of coded pulses that do not coincide with the moments of equality of the modulated signal and the reference constant; hereinafter, these fronts are called ayutsya stepped edges root coded pulses). Such pulses can be obtained either directly in the comparator, or on the basis of pulses obtained in the comparator.

Equation (1) and equality (2) mathematically describe the direct conversion of discrete values of an analog signal into an encoded form - the set t _r in an encoder, relation (3) the inverse transformation of an encoded form t _r into discrete values of a converted signal in a decoder. Consider the necessary conditions and requirements, the implementation of which will allow such a transformation.

As can be seen from (1) - (3), the modulating signal M (t) should not consist of rectangular pulses. Otherwise, the set t _r will contain randomly located roots. Obviously, to obtain from discrete values after converting a continuous analog signal as close as possible to the original analog signal, it is necessary that M (t) be a periodic signal. The higher the frequency of this signal, the closer the interpolated signal to the original signal. Within the period, the modulating pulse can be either single or in the form of a packet of pulses, however, it is clear that the use of a packet of pulses is simply equivalent to increasing the frequency of the modulating signal. The shape of the pulses can be quite diverse: symmetric or asymmetric with respect to the maximum of the pulse, with a stepped leading edge and monotonically decreasing from maximum to minimum trailing edge, with monotonously increasing from minimum to maximum leading edge and stepped trailing edge, etc. In practice, the shape of the pulse is chosen based on the greatest simplicity, reliability, stability, accuracy of electronic circuits generating a modulating signal in an encoder and a demodulating signal in a decoding device stve.

From (1), (2), (3) it follows that the working areas of the change of the analog and modulating signal should not include a zero point, because if any of the signals tends or is equal to zero, the other signal tends or is equal to infinity. Therefore, in the inventive method convert analog signals A (t), varying in the range of conversions
A _min ≅ A (t) ≅ A _max ,
where A _min > 0 and A _max, respectively, the lower and upper boundaries of the conversion range, and generate a modulating signal with a greater than zero lower boundary of the operating range.

Для уверенного срабатывания компаратора при A(t), близких или равных A_max и A_min, вводим значения верхней границы M_max рабочего диапазона и большей нуля нижней границы M_min рабочего диапазона такие, чтобы выполнялись неравенства:

из которых следует:
E <A_min•M_max, (6)
E > A_max•M_min. (7)
Объединяя (6) и (7), получаем
A_nax•M_min <A_min•M_max,
или
A_max/A_min <M_max/M_min. (8)
При выполнении неравенства (8) можно подобрать такую эталонную постоянную E, чтобы выполнялись неравенства (6) и (7). Для расширения области параметров модулирующих импульсов, которые можно использовать для преобразования, устанавливаем параметры модулирующих импульсов M_u ≥ M_max, M_l ≅ M_min, причем M₁ может быть равным или меньшим нуля. Наличие областей M(t) > M_max и M(t) < M_min не мешает преобразованию, поскольку в них корни уравнения (1) отсутствуют. Демодулирующий сигнал генерируют обратно пропорциональным по отношению к модулирующему. Скорее всего, наиболее рационально генерировать демодулирующий сигнал по следующей схеме. Сначала генерируют промежуточный сигнал, пропорциональный модулирующему, затем делят на него постоянную, такую, чтобы получить демодулирующий сигнал. При этом легко синхронизировать демодулирующий сигнал с временной шкалой кодированных импульсов путем синхронизации промежуточного сигнала.We determine the maximum M _u and the minimum M _{1 of the} modulating pulse. Using (1), we first find the minimum value of the maximum M $\genfrac{}{}{0ex}{}{\text{m}}{\text{u}}$ and the maximum value of the minimum M $\genfrac{}{}{0ex}{}{\text{m}}{\text{l}}$ modulating pulse:

For the comparator to operate reliably at A (t) close to or equal to A _max and A _min , we introduce the values of the upper boundary M _{max of the} working range and greater than zero of the lower boundary M _{min of the} working range such that the inequalities are satisfied:

from which it follows:
E <A _min • M _max , (6)
E> A _max • M _min . (7)
Combining (6) and (7), we obtain
A _nax • M _min <A _min • M _max ,
or
A _max / A _min <M _max / M _min . (eight)
If inequality (8) holds, one can choose a reference constant E such that inequalities (6) and (7) hold. To expand the range of parameters of modulating pulses that can be used for conversion, we set the parameters of modulating pulses M _u ≥ M _max , M _l ≅ M _min , and M ₁ can be equal to or less than zero. The presence of the regions M (t)> M _max and M (t) <M _min does not interfere with the transformation, since the roots of equation (1) are absent in them. A demodulating signal is generated inversely proportional to the modulating signal. Most likely, it is most rational to generate a demodulating signal according to the following scheme. First, an intermediate signal proportional to the modulating signal is generated, then a constant is divided into it, such as to obtain a demodulating signal. It is easy to synchronize the demodulating signal with the timeline of the encoded pulses by synchronizing the intermediate signal.

 In FIG. 1 shows graphical information illustrating the inventive method.

In FIG. 1a shows a modulating pulse. As an example, the function M (t) = cos ² (πt / T) + M _l is given within T / 2 <t <T / 2, for which M _max = M _u = 1.1, M _min M _l = 0.1 and M _max / M _min = 11. t _i and t _{i + 1 are the} starting and ending points of a period of duration T = t _{i + 1} -t _i .

 In FIG. 1b, an example of a fragment of an analog signal A (t) is given within one period, which varies from the beginning of the period to the end by 3 times, thereby inequality (8) holds.

 In FIG. Figure 1c shows the modulated signal A (t) • M (t) received in the comparator, in which it is compared with the reference constant E.

In FIG. Figure 1d shows the rectangular coded pulse obtained after the comparator, the leading and trailing stepped edges of which are root and their temporal coordinates t _r and t _{i + 1} are the moments of equality of the modulated signal and the reference constant.

 In FIG. 1e shows a rectangular pulse along with a synchronizing pulse.

 In FIG. 1e shows coded pulses in the form of short pulses obtained, for example, by differentiating a rectangular pulse, and a synchronizing pulse.

In FIG. 1g shows one period of the inverse function C / M (t). The values of C / M (t) at time t _r and t _{r + 1 are} proportional to A (t _r ) and A (t _{r + 1} ) of FIG. 1b.

 When using a modulating pulse with a stepped leading or trailing edge after the comparator, coded pulses are generated, for example, rectangular in which one of the stepped fronts coincides with the stepped edge of the modulating pulse. Obviously, the stepped edge of the encoded pulse, which coincides with the stepped edge of the modulating pulse, cannot be used to obtain the discrete value of the converted analog signal, since it is not the root. And if such a rectangular pulse is differentiated, then one of the two received pulses does not have a step front, which is the root front. A pulse that does not have a single step root edge is, by definition, not encoded and there is no need to transmit it to a decoding device.

The synchronizing pulses generated in the encoder together with the modulating signal are used to synchronize the time scales of the encoded pulses and the demodulating signal generator. In order that the synchronizing pulses do not coincide with the stepped root edges of the coded pulses, they are inserted at the points of the minimum or maximum of the modulating pulse. When inequalities (6) and (7) are fulfilled near the minimum and maximum of the modulating pulse, there are so-called “forbidden zones” for stepped root edges of coded pulses into which they do not fall. The durations of the “forbidden zones” are equal: in the vicinity of the minimum modulating pulse of time during which the modulating signal changes, decreasing from M $\genfrac{}{}{0ex}{}{\text{m}}{\text{l}}$ to M ₁ and then increasing to M $\genfrac{}{}{0ex}{}{\text{m}}{\text{l}}$ in the vicinity of the maximum modulating pulse of time during which the modulating signal changes, increasing from M $\genfrac{}{}{0ex}{}{\text{m}}{\text{u}}$ to M _u and then decreasing to M $\genfrac{}{}{0ex}{}{\text{m}}{\text{u}}$ If a modulating pulse with a stepped leading or trailing edge is used, the synchronizing pulse may coincide with such a front, however, this coincidence does not interfere with the conversion.

 The above graphic information shows that when using a modulating periodic pulse in the form of a monopulse with a monotonically increasing front and monotonically decreasing trailing edges, two discrete values of the analog signal are obtained in one period. When using a modulating pulse in which one of the fronts is stepped, one discrete value of the analog signal is obtained in one period. If you use a modulating pulse in the form of a pack of monopulses, then the number of discrete values increases accordingly.

The restoration of discrete values of the analog signal in the decoding device is carried out in various modes. In one of them, the demodulating signal generator operates continuously. At the moment of arrival of the stepped root edge of the encoded pulse, the value of the demodulating signal is recorded, for example, using a short selection pulse, and the result is sent to subsequent devices. If time is required to fix and transmit the discrete value of the analog signal, use the stop mode. For this, the started demodulating signal generator is stopped by the stepped root edge of the encoded pulse. The stopped value of the demodulating signal is recorded, the result is transmitted to subsequent devices, and the demodulating signal generator is prepared to restore the next discrete value of the analog signal. In those cases when the reconstructed discrete value is close in time to the points of the minimum or maximum of the modulating pulse, "forbidden zones" are used for the above operations. We denote the minimum required duration of the “forbidden zone” Δt _z . Obviously, from the point of view of the most rational use of the time factor, it is desirable to fulfill the inequality Δt _z <T. If, when using some form of a modulating impulse, inequalities (6) and (7) are insufficient to obtain the duration of the “forbidden zone” equal to Δt _z , use modulating pulse with a flat maximum and / or minimum, and the duration of the flat part is equal to Δt _z . In addition, recovery modes are possible using the delay of the pulses received in the encoder.

 Comparing the two methods of transmitting encoded information with each other - with rectangular or short pulses, obtained, for example, from a rectangular one by differentiating it, we note that from the point of view of minimal energy consumption by the encoding device (primarily its output stages), short pulses have an advantage, since the area the occupied by them is much smaller than the area of a rectangular pulse.

 Thus, in contrast to the prototype, in which the information on the amplitude of the analog signal is contained in a digital code or in the number of counting pulses, in the claimed method this information is enclosed in the time coordinate of the stepped root edge of the encoded pulse, which is a function of the shape of the modulating pulses for a certain value of the analog signal , and for the selected form of the modulating pulse by a continuous function of the analog signal. In view of this, the claimed method can be called a functional-temporary converter.

 Here are the comparative characteristics of the prototype and the proposed method.

 1. Informational content.

 In the prototype for one clock period receive one discrete value of the analog signal.

 In the inventive method, when using a modulating pulse in the form of monotonically changing leading and trailing edges, two discrete values of the analog signal are obtained in one period.

 2. The accuracy of the conversion.

 In the prototype, the converted value is undefined within the same quantization level both in amplitude and in time.

 In the inventive method, an exact discrete value is obtained with an exact time reference.

 3. The number of pulses transmitted to the decoding device per one discrete value of the analog signal, not counting the synchronizing pulses.

 In the prototype, the number of counting pulses within one clock period is large and it grows as the accuracy of quantization increases.

 In the inventive method, one discrete value is transmitted in one short pulse.

 4. The complexity of electronic devices.

 The prototype uses complex digital circuits in both encoding and decoding devices.

 In the inventive method can be used quite simple devices for multiplying, dividing signals and generating periodic pulses of non-rectangular shape.

 5. Interference immunity when transmitting encoded information over communication lines or over the air.

 The noise immunity of the prototype against amplitude distortion is known. The noise immunity of the proposed method is ensured by the fact that the amplitude distortion does not shift in time the stepped root front of the encoded pulse.

 The comparison indicates the significant advantages of the proposed method.

 The possibility of implementing the proposed method is not in doubt, since the operations of multiplication and division of signals, as well as the generation of periodic pulses of non-rectangular shape in electronics have long been known and are widely used. The accuracy of the conversion depends on the accuracy of the generation in the decoding device of a demodulating signal inversely proportional to the modulating.

Literature
1. Svoren R. Electronics of communication, w / l. Science and Life, N 10, p. 2-11, tabs II-III, 1986.

 2. Conversion of information in analog-to-digital computing devices. Ed. G.M. Petrova. M. Engineering, 1973.

 3. B.V. Anisimov, V.N. Chetverikov. Conversion of information for digital computers. M. Higher School, 1968.

Claims (1)

A method of converting an analog signal, which consists in generating synchronizing pulses, in accordance with which periodic time intervals are formed, and in each of them a discrete value of the analog signal is encoded by generating encoded pulses, which, when converted inverse, decode to a discrete value of the analog signal, different by converting the analog signal A (t), varying in the conversion range
A _{m i n} ≅ A (t) ≅ A _{m a x} ,
where A _{m i n} > 0 and A _{m a x} are the lower and upper boundaries of the conversion range, in each periodic time interval, a modulating signal is generated in the form of non-rectangular pulses, the minimum M _L and maximum M _U values of which satisfy the inequalities
M _L ≅ M _{m i n} , M _U ≥ M _{m a x} ,
where M _{m i n} > 0 and M _{m a x} are the lower and upper boundaries of the working range, respectively
A _{m a x} / A _{m i n} <M _{m a x} / M _{m i n} ,
multiply the analog signal by a modulating signal and the resulting modulated signal is compared with a reference constant E, the value of which satisfies the inequality
A _{m a x} M _{m i n} ≅ E <A _{m i n} M _{m a x} ,
form coded pulses whose edges coincide with the moments of equality of the modulated signal and the reference constant E, when the inverse transform is generated, in accordance with the synchronizing pulses, a demodulated signal inversely proportional to the modulating signal, and decoded the encoded pulses into discrete values of the analog signal by determining the values of the demodulating signal at times fronts of coded pulses coinciding with the moments of equality of the modulated signal and the reference signal Toyan E.