RU2251118C2 - Non-linear distortions evaluation calculator - Google Patents

Abstract

FIELD: electric measurement technology.

SUBSTANCE: device is used for evaluating presence and degree of non-linear distortions of real random signal when the signal passes sound-amplifying section. Evaluation of non-linear distortions is based upon measurement of dispersion of output-to-input signal ratio at small group time lag. Distortions can be evaluated without interrupting operation of amplifier or switching it to specific mode of operation.

EFFECT: improved precision of measurement; improved reliability of operation.

3 cl, 1 dwg

Description

The invention relates to the field of electro-radio measurements and is intended to assess the presence and degree of non-linear distortion of a real random signal when it passes through sound amplifying paths.

The prototype of the claimed device is a structure containing a test signal generator, a notch filter, two rms meters and a division unit, the output of which is the information output of the device, the test output of which is the output of the test signal generator, and the test input is the input of the notch filter, the output of which is connected to the input of the first rms meter, the output of which is connected to the first input of the division unit, the second input of which is connected to the output of W cerned RMS meter having an input merged with the input of the notch filter [Meyzda F. Electronic Instrumentation and Measurement methods: Trans. from English -M.: Mir, 1990, p. 384, Fig. 14.14b].

The principle of operation of the prototype involves the use of a special measuring signal, usually monoharmonic, according to the results of distortion of which they judge the presence and degree of non-linearity of the path. This feature does not allow the use of devices of this type to evaluate non-linear distortions arising in the process of the amplifier performing its regular functions - amplification of a useful signal, which in reality is random. Marked refers to a significant disadvantage of the prototype.

The technical result achieved by using the present invention is to provide the ability to identify and evaluate non-linear distortions in the process of the amplifier performing its useful functions, without interrupting its operation and transferring it to a special measuring mode.

The technical result is achieved by the fact that in the non-linear distortion estimation calculator containing the division unit, according to the invention, a dispersion estimation unit is introduced into it, the output of which is the output of the calculator, the first and second test inputs of which are the first and second inputs of the division unit, the output of which is connected to the input of the variance estimation block.

According to the second embodiment, the technical result is achieved in that, in accordance with the invention, two band-pass filters and a dispersion estimation unit are introduced into the non-linear distortion estimation calculator containing the division unit, the output of which is the output of the calculator, the first and second test inputs of which are the inputs of the first and second band filters the outputs of which are connected respectively to the first and second inputs of the division block, the output of which is connected to the input of the dispersion estimation block.

The invention is illustrated graphic material. The drawing shows a functional diagram of a computer with a connected tested amplifier.

The functional diagram of the calculator for estimating nonlinear distortions contains a division unit 1, a dispersion estimation unit 2, and a test amplifier 3 with a load resistance R _L. The first test input of the calculator, which is the first input of the division unit 1, is combined with the input x (t) of the tested amplifier 3, the output y (t) of which is connected to the second test input of the calculator, which is the second input of the division unit 1, the output of which is connected to the input of the unit 2 estimates of variance, the output of which serves as the output of the computer - the output of the estimate of non-linear distortions.

The process of calculating the desired estimate of nonlinearities is as follows.

The inputs of the division unit 1 receive the input x (t) - the divider and the output y (t) - the dividend signals of the tested amplifier 3, which can be either random or deterministic - the principle of the device does not change from this. In the absence of nonlinear distortions and a negligible group delay time, the connection between the output and input of amplifier 3 is unambiguously described by the simple expression y (t) = kx (t) (k is the proportionality coefficient, which is physically the gain in the linear region). If we assume that during the chosen interval for observing the operation of amplifier 3, its parameters do not change, and the values of the input signal x (t) are within the linear gain, that is, x ₁ ≤ x (t) ≤ x ₂ ([x ₁ ; x ₂ ] is the linear gain range), then at the output of the division unit 1 there will be a constant voltage equal to the value of k. When non-linear distortions appear, for example, due to the output of the signal x (t) outside the range [x ₁ ; x ₂ ], the relationship between the output and input signals will be expressed through the functional

which is also a random variable for a random input signal x (t).

To assess the presence and degree of nonlinear distortion, one should determine the variance of the above functional, i.e., the value

дисперсия

так как результат деления k есть постоянная величина - это случай полного отсутствия нелинейных искажений: на выходе блока 2 напряжение равно нулю. Появление нелинейных искажений приводит к появлению функционала F(x(t)), зависящего от характера нелинейностей: чем ярче выражена нелинейность, тем в большем диапазоне изменяется F(x(t)), а следовательно, и больше вычисленное значение дисперсии на выходе блока 2, разумеется уже отличное от нуля.The calculation of the variance of the result of dividing the output signal by the input allows one to statistically evaluate how much the gain of the signals varies depending on their instantaneous values, and therefore, how non-linear the transfer characteristic of the studied amplifier for signals with given amplitude values. Thus, by the value of the value at the output of block 2 of the variance estimate, non-linear distortions should be judged. At

dispersion

since the result of the division of k is a constant value, this is the case of the complete absence of nonlinear distortions: at the output of block 2, the voltage is zero. The appearance of nonlinear distortions leads to the appearance of the functional F (x (t)), which depends on the nature of the nonlinearities: the more pronounced the nonlinearity, the larger the range of F (x (t)), and therefore, the greater the calculated dispersion value at the output of block 2 , of course, is already non-zero.

возможно, придется принимать не за нуль, а за некоторый, отличный от нуля, пороговый уровень, появление которого вызвано указанными выше факторами, но это также зависит от требуемой точности измерений и при определенном загрублении результатов, при невысоком разрешении отсчет может действительно начинаться с нуля.The estimation errors in this calculator will largely depend on finding the ratio y (f) / x (t), where the reaction y (t) at each moment of time must strictly correspond to the effect x (t) that caused it. Such a correspondence is possible only in the absence of any delay in the transmission of influence from input to output. Of course, this does not happen in reality, and the output signal is always late relative to the input. However, in low-cascade low-frequency amplifiers, the group delay time is so short that it can be neglected in many problems. (Strictly speaking, the correlation interval of amplified signals in low-frequency amplifiers is much larger than the group delay time.) At the same time, it is useful to theoretically take into account the presence of delay in each specific case, as well as the presence of noise in the output signal y (t). Therefore, in practice, the lower limit of the quantity

it may be necessary to take not a zero, but a certain threshold level other than zero, the appearance of which is caused by the above factors, but it also depends on the required measurement accuracy and, with a certain coarsening of the results, with a low resolution, the reading can really start from zero.

Since a real random signal with a continuous spectrum can be used as a measuring signal, one should take into account the possibility of uneven (unequal) amplification of all frequency components, which, in turn, can lead to another component of the error. In the event that the non-uniformity of the amplitude-frequency characteristic can really affect the result, it is advisable to apply signals to the inputs of block 1 in a limited frequency band, the unevenness of the frequency response of the amplifier, within which it can be neglected (see the dependent claim).

Claims (2)

1. A nonlinear distortion estimation calculator comprising a division unit, characterized in that a dispersion estimation unit is introduced into it, the output of which is the output of the calculator, the first and second test inputs of which are the first and second inputs of the division unit, the output of which is connected to the input of the evaluation unit variance. 2. A nonlinear distortion estimation calculator comprising a division unit, characterized in that two band-pass filters and a dispersion estimation unit are introduced into it, the output of which is the output of the calculator, the first and second test inputs of which are the first and second inputs of the band-pass filters, the outputs of which are connected respectively, with the first and second inputs of the division block, the output of which is connected to the input of the dispersion estimation block.