RU2520956C2 - Digital meter of amplitude frequency response - Google Patents

Abstract

FIELD: measurement equipment.

SUBSTANCE: proposed invention relates to measuring equipment, in particular, to amplitude frequency response measurement. A digital amplitude frequency response meter comprises an indicator, a microcontroller, an analogue-to-code converter with its output being connected to the microcontroller input. The first microcontroller output is connected to the indicator. The device also comprises a code-to-analogue converter, an input device and an interface device with its first input being the meter input and the first output being the meter output. The second output of the interface device is connected to the first input of the analogue-to-code converter with the second output of the latter being connected to the second output of the microcontroller. The second microcontroller input is connected to the input device. The third microcontroller output is connected to the third input of the interface device, the fourth microcontroller output is connected to the input of the code-to-analogue converter with the output of the latter being connected to the second input of the interface device.

EFFECT: expanded functionality due to the possibility of AFR measurement.

5 dwg

Description

The present invention relates to the field of measuring equipment, in particular to means for measuring the amplitude-frequency characteristics (AFC) of a four-terminal network.

A device for automatically measuring the parameters of the amplitude-frequency characteristics of a selective quadrupole having only one maximum or minimum in its frequency response [A. S. USSR No. 375588, IPC G01R 27/28, Publ. 1973], containing a triangular voltage generator, oscillating frequency generator, tested quadrupole, detector, threshold pulse shaper, counting interval shaper (FIS), a reversible counter and a digital indicator that provides an indication of the measurement results. Improving the accuracy in comparison with the prototype is achieved by compensating for the methodological and dynamic measurement errors due to the fact that the total counting time of the unknown frequency is divided into two parts, one of which has a time shift up and the other down with respect to the true frequency value. However, the device cannot work correctly under the condition of several maxima in the studied frequency response, and also display the measured frequency response on the indicator device.

Closest to the technical nature of the claimed device is selected as a prototype digital modulation meter [Pat. 2424534 RF, IPC G01R 29/06, Publ. 07/20/2011], containing an indicator, an analog-code converter, an input device and a microcontroller.

The block diagram is shown in figure 1. The digital meter contains an input device 1, an analog-code converter 2, a microcontroller 3, and an indicator 4. Moreover, the output of an input device 1 is connected to an analog-code converter 2 an input, the output of which is connected to the first input of the microcontroller 3, the first output of the microcontroller is connected to an indicator, the second output is connected to the second input of the converter 2 analog code, and the third output is connected to the second input of the input device 1.

The principle of operation of the meter is based on processing a discretized data array using Fourier and Hilbert transforms.

The Hilbert Transform allows it to find an orthogonal signals X ₁ (t) to X (t) signal. Using these signals, the envelope (instantaneous amplitude) and the instantaneous frequency of the signal are found by the formulas:

;

.

;

.

Due to additional processing and filtering, the following derived parameters are found:

- Depth AM - peak and root mean square values in a given frequency band.

- Frequency deviation - peak and rms value in a given frequency band.

- Carrier frequency (center frequency).

- The frequency of the modulating signal AM and (or) FM.

- The coefficient of nonlinear distortion of the modulating signal AM and (or) FM.

The device does not have a signal generation circuit and, therefore, cannot be used to measure the amplitude-frequency characteristics.

The object of the invention is to expand the functional capabilities of the device, namely providing the possibility of measuring the amplitude-frequency characteristics (frequency response) of quadripole, such as: the lower frequency at a given level -f _n, upper frequency -f at a predetermined level _in the frequency band at a predetermined level f _in -f _n , the center frequency at a given level f _c = (f _in -f _n ) / 2, the frequency response unevenness in a given frequency band - α.

The task is achieved in that a digital modulation meter containing an indicator, a microcontroller, an analog-to-code converter, the output of which is connected to the first input of the microcontroller, the first output of the microcontroller is connected to an indicator, a code-to-analog converter, an input device and a pairing device, the first input which is the input of the meter, and the first output is the output of the meter, the second output of the interface device is connected to the first input of the analog-code converter, the second input of which is connected ene to the second output of the microcontroller, a second input coupled to the input device, the third output of the microcontroller is connected to the third input of the coupling device, the fourth output of the microcontroller is connected to the input of the converter digital-to-analog output of which is connected to the second input coupling device.

The structural diagram of a digital frequency response meter is shown in figure 2. On the diagram are indicated: interface device 1, converter 2 analog-to-code, microcontroller 3, indicator 4, converter code-analog 5, input device 6. Moreover, the first input of the interface 1 is the input of the meter, and the first output is the output of the meter, the second output of the device interface 1 is connected to the first input of the converter 2 analog code, the second input of which is connected to the second output of the microcontroller 3, the second input of which is connected to the input device 6, the third output of the microcontroller 3 is connected to the third input of the device interface 1, the fourth output of the microcontroller 3 is connected to the input of the digital-analog 5 converter, the output of which is connected to the second input of the interface device 1, the output of the converter 2 analog-code is connected to the first input of the microcontroller 3, the first output of the microcontroller 3 is connected to the indicator 4.

The device operates as follows.

The signal from the studied device or communication line is fed to the interface device 1, which is a matched amplifier-attenuator with an adjustable transmission coefficient for both the receiving and transmitting paths. Then the signal goes to the converter 2 analog code (analog-to-digital converter - ADC), operating in the gating mode. Depending on the level of the signal digitized by the analog-to-analog converter 2, the microcontroller 3 sets the transfer coefficient of the receiving path of the interface device 1 so that the maximum value of the ADC code lies in the range from 0.5 to 0.9 of the scale limit. Depending on the operating mode selected through input devices 6 and displayed on indicator 4, microcontroller 3 sets the sampling frequency f _∂ for analog-to-code 2 and analog-to-code 5 converters so that the spectrum of the generated and received signal is in the frequency range from 0 to f _∂ / 2. An array of digitized data X [iT _∂ ] from the output of the converter 2, the analog code is sent to the microcontroller 3, which implements the algorithm for operation presented below.

Depending on the selected operating mode, the device displays on indicator 4 either the frequency response of the measured path with the calculated frequency response parameters, or the parameters of the modulated signal.

The principle of operation of the meter is based on the advanced processing of the sampled data array using the Fourier and Hilbert transforms.

The following algorithm is implemented in the microcontroller:

If the “Digital modulation meter” operating mode is selected, the operation algorithm completely repeats the operation algorithm of the device from the prototype.

If the “Digital frequency response meter” operating mode is selected, a modified operating algorithm is implemented that uses the digital method of compensating dynamic measurement errors due to the fact that the total counting time of the unknown frequency is divided into two parts, one of which has a time shift up, and the other down to the true value of the frequency, created on the basis of the analogue method.

Work algorithm:

1. Select the sampling frequency f _s . According to Comrade Kotelnikov, f _s should be more than two times greater than the upper frequency in the spectrum of the analyzed signal. Additionally, it is necessary to take into account the expansion of the spectrum due to the need to use frequency modulation (FM) of the test signal, the minimum expansion occurs when using harmonic frequency modulation (GFM). A filter is used in the interface device to suppress the frequency components f _s / 2, however, the level of suppression of these components may be low, for example 12 dB per octave. T.O. to use the same sampling frequency for the DAC and ADC, it is advisable to choose a sampling frequency 4 times greater than the upper frequency in the spectrum of the test FM signal, thereby weakening the filter requirements in the interface device. According to T. Kotelnikov, the sampling frequency should be more than 2 times higher than the upper frequency of the spectrum of the generated signal.

2. We generate a test GFM signal with a sampling frequency f _s containing 1 modulation period, i.e. frequency changes from f _min to f _max , from f _max to f _min . In the microcontroller’s memory we store an array of values u _tst [iT _∂ ], where i∈ [0, N-1] is the number of the element in the array u _tst [iT _∂ ], consisting of N points.

Where U _max - the amplitude of the generated DAC signal in discrete, for example, 5123,

Δ _f is the maximum frequency deviation relative to the average Δ _f = (f _max -f _min ) / 2,

f _max and f _min - user-defined values of the frequency range,

f _mod = f _s / N is the modulation frequency,

- центральная частота.

- center frequency.

This choice of signal frequency allows you to specify exactly one period of the test signal in the array.

3. _{We feed} an array of values u _tst [i] to the input of the converter.

4. We sample the output signal of the tested module with a frequency f _s and obtain an array u [i], where i∈ [0, N-1] is the number of the element in the array u [i].

5. Find the maximum value from the array A _MAX = MAX (u [i]).

If P ₁ > A _max / MAX _ADC > P ₂ , then the input gain of the input device is not changeable. Here: P ₁ and P ₂ are the maximum and minimum coefficients of using the dynamic range of the ADC (you can choose P ₁ = 0.9, P ₂ = 0.5); MAX _ADC - ADC scale limit. If necessary, change the transmission coefficient, repeat the 4th and 5th steps of the algorithm. The right choice of transmission coefficient will provide more complete use of the ADC working range, which will contribute to high digitization accuracy.

6. We find the direct fast Fourier transform (FFT) from the array u [i], we obtain an array of spectral components S [i] = FFT (u [i]). To filter the parasitic constant component that occurs in the process of obtaining a discretized data array, zero the amplitude of the 0-th spectral component and get the array S ^* [i]. Then, using the inverse Fourier transform (OBPF), we get the filtered array u ^* [i].

7. We find the Hilbert transform from the array u [iT _∂ ] via FFT and FFT (RFT):

u _⊥ [i] = H (u ^* [i]) = RFT (k · S _⊥ [i]), where S _⊥ [i) = k · S ^* [i]

k = -j if i = 0, 1, 2, 3, ... N / 2; k = j if i = N / 2 + 1, N / 2 + 2, N / 2 + 3, ... N-1.

.8. We find the envelope of the frequency response by the formula:

.

9. Zero the high-frequency spectral components of the envelope of the frequency response. To do this, we calculate the direct Fourier transform of the array A [i] (An example of the array A [i] is shown in Figure 3). In the resulting array of spectral components S _A [i], we zero out the components from Q to NQ, where Q is a user-defined parameter, for example 64. Parameter Q essentially sets the maximum number of periods (maxima and minima) of the frequency response after filtering. The smaller this parameter, the stronger the frequency response is filtered, but the more gentle the frequency response will be, i.e. maximum slope is determined by this parameter. The real value of Q can vary in the range from 16 to N / 2, with N / 2 filtering is completely disabled. Then we calculate the inverse Fourier transform, the results of which are written into the array A ^* [i]. This operation will eliminate the ruggedness of the amplitude-frequency characteristics caused by noise. An example of an array A ^* [i] is shown in Figure 4.

10. In order to find the instantaneous frequency, we calculate the derivative of the arrays u ^* _, u _⊥ using the direct and inverse Fourier transform:

(u [i]) ^′ = RFT (k · S ^* [i]); (u _⊥ [i] ^′ = RFT (k · S _⊥ [i]).

Here k = jωi if i = 0, 1, 2, 3, ... N / 2; k = -jω (N-i) if i = N / 2 + 1, N / 2 + 2, N-1.

11. We get an array of instantaneous frequency values F [iT _∂ )]:

.

.

. Находим обратное преобразование Фурье от массива

, получаем отфильтрованный массив F^*[i].12. We find the Fourier transform of the array F [i], we obtain the array S _F [i] = FFT (F [i]). We filter out the array S _F [i], zeroing out the components with indices i = 3, 4, ... N-3, we get an array

. Find the inverse Fourier transform of the array

, we get the filtered array F ^* [i].

13. Connect the arrays of instantaneous values F ^* [i] and A ^* [i] into a single array of values FA [i] .F, FA [i] .A. Also, for each element of the array, add the binary parameter I, which takes the values true / false (1 and 0, respectively). The value "true" means that the point belongs to the increasing section of the amplitude, the value "false" means that the point belongs to the decreasing section of the amplitude. When combined, the parameter I = 0. Thus, the FA array is a table of records consisting of three fields:

- Amplitude, A.

- Frequency, F.

- Parameter of increase, I.

14. We divide the array FA [i] into two arrays, in one of which with increasing index the frequency increases - F ^↑ A [i], in the other it decreases F ^↓ A [i] accordingly. To do this, perform steps 14.1.-14.4.

14.1. We set the value of the indices of arrays i = 0, j = 0, k = 0.

14.2. If FA [(i + 1)]. F> FA [i]. F> 0, we put the point FA [i] in the array F ^↑ A [j] (see Figure 5, the dash-dotted line) and increase the value of j by 1, otherwise we enter this point in the array F ^↓ A [k] (see 5, dashed line) and increase the value of k by 1.

14.3. Increase i by 1.

14.4. If i <N-1, go to step 14.2.

15. We calculate for arrays F ^↑ A, F ^↓ A the parameter of increase is I.

16. Set the index of the array i = 0.

17. We calculate the difference in the values of the amplitudes as dA = F ^↓ A [i + 1] - F ^↓ A [i]

18. If dA> 0, then for the point F ^↓ A [i + 1] we establish the sign of increasing I = 1. If i = 0, then the same operation is performed with the point F ^↓ A [0], and the sign I for this point is defined as I for the point with index 1.

19. Increase the value of the array index by 1. If i <N, go to step 18.

20. Repeat paragraphs of algorithm 16-19 for an array F ^↑ A.

21. Set the index of the array i = 0. The value of the attribute of the previous point U _{Aprev is} set to 0. The value of the sort boundary Bord = 0.

22. Set t equal to Bord, ΔА equal to the maximum possible value of amplitude A.

23. Compare the point F ^↑ A [i] with the point F ^↓ A [t] in amplitude. We calculate the difference in values and, if it is less than ΔА, then ΔА we assign its value, and also calculate the average value of the amplitude

mA = (F ^↑ A [i]. A + F ^↓ A [t] .A) / 2

24. Increase t by one.

25. If for the point FA ^, [t] I _A = I _Aprev , go to step 24. Otherwise, we write the value of the amplitude mA into the array FA _sort [i] .A, FA - _sort [i] .A = mA, calculate the average value frequency

mF == (F ^↑ A [i] .F + F ^↓ A [t] .F) / 2

and also write the value to the array FA _sort [i] .F = mF. Increase i by one and go to step 27.

26. If for an array point FA '' [i] I _A ≠ I _{Aprev, we} assign the value of the sort boundary Bord = i. If i = N, go to step 28; otherwise, repeat the steps from step 23.

27. If the relative measurement mode is not selected, then we display the obtained dependence FA _sort [i] (see Figure 5, solid line) on the indicator device.

28. We calculate and display the non-uniformity - α in the frequency range from f _min to f _max specified by the user. α = 201g (A _max / A _min ), where A _max and A _min are the maximum and minimum values in the array FA _sort [i] lying between the frequencies f _min and f _max .

29. Calculate the maximum value of the amplitude A * in the array FA _sort [i].

на индикаторном устройстве.30. We normalize the array FA _sort [i] according to the maximum amplitude FA ^* _sort [i] .A = 20lg (FA _sort [i] .A / A ^* ). If the relative measurement mode is selected, then we display the obtained dependence

on the indicator device.

находим ближайшие к L значения амплитуды

. Определяем, какое из найденных значений массива больше по частоте, записываем частоту для этого индекса в переменную

, меньшее значение по частоте записываем в переменную f_н.31. We determine the lower -f _n and the upper -f _{to the} frequency at the user-specified level L [dB], for this in the array

we find the amplitude values closest to L

. We determine which of the found values of the array is larger in frequency, write the frequency for this index into a variable

, a lower value in frequency is written into the variable f _n .

32. We calculate the frequency band at a given level f _in -f _n and the center frequency at a given level F _c = (f _in -f _n ) / 2.

33. We display the found values of the frequency response parameters: α, f _in -f _n , f _in , f _n , f _c .

34. If analysis requires changing the frequency range for analysis, then go to step 2 of the algorithm, otherwise go to step 3.

Paragraphs of algorithm 13-26 implement the compensation of dynamic measurement errors due to the fact that the frequency response measurement is divided into two parts, one of which occurs with increasing frequency, and the other with decreasing frequency.

The greatest effect of using the proposed invention can be achieved in measuring complexes containing a high-speed microcontroller / signal processor. The expansion of functionality was achieved by complicating the digital processing algorithm and introducing a test signal generation circuit for measuring the frequency response.

The proposed digital frequency response meter can measure:

- Frequency response of the device.

- Frequency response of a communication line or channel, while 2 meters are needed at each end of the communication line or channel.

- Frequency response parameters: lower -f _n and upper -f _to frequency at a given level, frequency band at a given level f _to -f _n , center frequency at a given level f _c = (f _to -f _n ) / 2, frequency response unevenness given frequency band - α.

- Depth AM - peak value in a given frequency band.

- Frequency deviation - peak value in a given frequency band.

- Depth AM - RMS value in a given frequency band.

- Frequency deviation - RMS value in a given frequency band.

- Carrier frequency (center frequency).

- The frequency of the modulating signal AM and (or) FM.

- The coefficient of nonlinear distortion of the modulating signal AM and (or) FM.

The use of an inexpensive digital circuitry base in a frequency response meter reduces the cost and improves the reliability of the device.

Claims (1)

A digital amplitude-frequency characteristics meter containing an indicator, a microcontroller, an analog-to-code converter, the output of which is connected to the first input of the microcontroller, the first output of the microcontroller is connected to an indicator, characterized in that a code-to-analog converter, an input device and a pairing device are inserted into it, whose first input is the input of the meter, and the first output is the output of the meter, the second output of the interface device is connected to the first input of the analog-code converter, the second input is cerned is connected to the second output of the microcontroller, a second input coupled to the input device, the third output of the microcontroller is connected to the third input of the coupling device, the fourth output of the microcontroller is connected to the input of the code-to-analog converter whose output is connected to the second input coupling device.

Images