JPH04299269A - Measuring apparatus for wide-band effective value - Google Patents

Abstract

PURPOSE:To provide the title apparatus accurately measuring (reproducing) a signal to be measured even when the sampling frequency of an A/D converter is twice or less that of the signal to be measured. CONSTITUTION:An apparatus for measuring a wide-band effective value is constituted of an A/D converter means (2) sampling a signal to be measured at a plurality of different fixed cycles, a square operation means (4) individually squaring sampling data at the respective fixed cycles of the A/D converter means (2), a digitalizing average operation means (5) individually performing digitalizing average operation with respect to the square operation data of the square operation means 4, an extraction-of-square root means 6 individually performing extraction of the square root with respect to the digitalizing average data of the digitalizing average operation means 5 and an operation means (7) discriminating and removing the max. and min. values from the extraction-of- square root data of the means 6 to operate the average of the residual data.

Description

[Detailed description of the invention]

0001

[Field of Industrial Application] The present invention relates to a wideband effective value measuring device, and more particularly, to a digital sampling type effective value measuring device that digitally samples a signal to be measured (input voltage or input current) and measures its effective value. Regarding broadbandization.

0002

[Prior Art] Conventionally, as an effective value measuring device using a digital sampling method, a signal under test is applied to an A/D converter, repeatedly sampled at a cycle shorter than the cycle of the signal under test, and the output data is processed. In addition to the circuit,
The data processing circuit used is one that multiplies output data and calculates the average value of the multiplication results.

That is, if the sampling data of the signal under test B is Bn, its effective value BRMS is:

000
4]

[Math 1]

It is expressed as 0005. Here, if the signal to be measured B is a voltage signal, Bn is a sampling voltage value, and if it is a current signal, Bn is a sampling current value.

By the way, according to such a conventional configuration, if the sampling frequency of the A/D converter becomes less than twice the frequency of the signal under test, the signal under test cannot be accurately measured (reproduced) based on the sampling theorem. It's gone.

[0007] Therefore, in order to measure high-frequency signals, it is necessary to use a high-speed A/D converter capable of high-speed sampling.

[0008]

[Problem to be solved by the invention] However, high speed A
A/D converters are expensive, and A/D converters currently sold in the general market have a limited sampling band, although they have a wide band.

The present invention has been made in view of these problems, and its purpose is to accurately measure the signal under test even if the sampling frequency of the A/D converter is less than twice that of the signal under test. The object of the present invention is to provide a wideband effective value measuring device that can measure (reproduce) well.

0010

[Means for Solving the Problems] The present invention solves the above problems by providing an A/D conversion means for sampling a signal under measurement at a plurality of different fixed periods, and an A/D conversion means for sampling a signal under measurement at a plurality of different fixed periods. squaring calculation means for individually squaring the sampling data of the squaring calculation means; exponentiated average calculation means for individually performing an exponential averaging calculation on the squared calculation data of the squaring calculation means; Consisting of a square root calculation means that individually performs a square root calculation on the exponential average data, and a calculation means that determines and removes the maximum value and minimum value from the square root data of the square root calculation means and calculates the average of the remaining data. It is characterized by the fact that

[0011]

[Operation] The A/D conversion means converts the analog measured signal (B)
The instantaneous value (Bn) of multiple different fixed periods CLK (1)
~CLK(n) and converted to digital data.

The squaring calculation means individually squares the sampling data at each fixed period CLK(1) to CLK(n) of the A/D conversion means, thereby converting the analog signal under measurement (
The square value (Bn2) of each instantaneous value (Bn) of B) is output.

[0013] The indexed average calculation means individually performs the indexed average calculation a predetermined number of times on each squared calculation output data (Bn2) of the squared calculation means. As a result, the indexed average value of (Bn2), that is,

[0014]

[Math 2]

[0015] is obtained.

The square root calculation means performs a square root calculation on the indexed average of (Bn2), thereby obtaining the square root of the indexed average of (Bn2), that is,

[0017]

[Math 3]

Outputs . The square root data is A/
For the fixed cycle CLK(1) of the D conversion means, S(1)
is output, and for fixed period CLK(n), S(n)
will be output.

The calculation means uses these square root data S(1)
~S(n) is introduced. And these square root data S
(1) Determine and remove the maximum and minimum values from S(n) and calculate the average of the remaining data. For example, S(1)~
Among the six data of S(6), S(2) is the maximum value and S
If (5) is the minimum value, then

[0020]

[Math 4]

The following calculation is performed.

[0022] Thereby, the true effective value of the signal under measurement can be obtained.

[0023]

Embodiments Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram of an embodiment of the present invention, and FIG. 2 is a signal waveform diagram of each part of an arbitrary system (for example, the system with subscript 1) in FIG.

In FIG. 1, 1 indicates the signal under test B (FIG. 2(
1)), and the input terminal 1 is an input terminal for a plurality of A/D converters 21 to 2 including sample and hold circuits.
connected to n. Clock generators 31 to 3n supply sample clocks with different fixed periods CLK(1) to CLK(n) to the A/D converters 31 to 3n. Each A/D converter 21 to 2n has a fixed period CLK to which the instantaneous value of the analog signal under test B is added to each one.
(1) - Sampling at CLK(n) and converting to digital data. For example, every time a clock is applied from the clock generator 31, the A/D converter 21 converts the instantaneous value B(t) of the signal under measurement B at that time into a digital signal. The black circles in FIG. 2(1) indicate each sampling point. t means the number of sampling times, t=1, 2,
3,..., and B(1)=B1, B(2)=B2,...,
It is expressed as B(n)=Bn. That is, at the n-th sampling, the instantaneous value Bn is output. Note that TS in FIG. 2 indicates the sampling start time, and TAD indicates the sampling period.

41 to 4n are square calculation units, each A/D
Data obtained by individually squaring the sampling data of the converters 21 to 2n is output. For example, the squaring calculator 41 uses t
= Square calculation data of instantaneous value Bn at n (n) = B
Output n2. The output of the square calculator 41 is digital, but when expressed in analog form, it becomes a solid waveform as shown in FIG. 2(2). Since the solid line waveform is the square value of the sampling data B1, B2, B3, . . . , Bn, . . ., it has a step-like shape. Note that the dotted line waveform in Figure 2 (2) is similar to Figure 2 (1).
) is an ideal square waveform of the waveform.

Reference numerals 51 to 5n represent exponential averaging units, which individually exponentially average the square calculation data of each of the square computing units 41 to 4n. For example, if the square calculation data b(n) output from the square calculation unit 41 is exponentially averaged by the exponential averaging unit 51 and the output P(n) is expressed in analog form, FIG.
(3) becomes. That is, the exponential averaging calculator 51 includes an arithmetic unit and a memory. The calculation section repeats the calculation of the following equation every time the square calculation output data b(n) is added from the square calculation unit 41.

P(n)=P(n-1)+(1/G)·{b(n)
-P(n-1)}...(5) P(n): Indexed average result processed up to sampling number t=n P(n-1): Processed up to sampling number t=n-1 Exponential average result b(n): Square value of the instantaneous value of the measured signal at sampling number t=n 1/G: Exponential constant ((1/G)<<1) where, (
1/G)<<1, so if the calculation of equation (5) is repeated every time it is added to the exponential averaging calculator 51, the result shown in FIG.
The waveform becomes 3). That is, equation (5) means that by repeating this calculation, P(n) becomes the average value of the waveform of b(n). Therefore, the final value in FIG. 2(3) coincides with the average value of b(n). In other words, from the exponential averaging calculator 51, the above-mentioned (
2) The value of the expression can be obtained. Note that although Figure 2 (3) is drawn so that the value of P(n) becomes constant after a small number of operations, in reality, for example, the value of P(n) can be changed by repeating the number of operations n = 1000. becomes constant.

61 to 6n are respective exponential averaging calculation units 51 to 5.
This is a square root calculator that individually performs a square root calculation on n indexed average data. For example, the square root calculation unit 61 is the exponential average calculation unit 5.
The square root of the data in equation (2) above that is added from 1 is calculated, and the square root data expressed in equation (3) above is output. As is clear from equation (1) above, this represents the effective value of the signal under measurement B.

[0030] The square arithmetic unit, the exponential average arithmetic unit, and the square root arithmetic unit can be constructed using, for example, a digital signal processor (DSP). According to such a configuration, the measurement data D(1) to D(n
) can obtain simultaneity.

[0031] 7 is an arithmetic unit (CPU), as shown in Fig. 2 (3).
P(
The maximum value and the minimum value are discriminated and removed from the square root data of each of the square root calculation units 61 to 6n regarding n), and the average of the remaining data is calculated, and the calculation result as shown in equation (4) above is displayed as an effective value on the display 8. Output to.

The reason for the arithmetic processing in the arithmetic unit 7 is shown in FIG.
Explain using. FIG. 3 is an explanatory diagram of the relationship between the sampling theorem and measured values.

When the signal frequency to be measured is fin and the sampling frequency of the A/D converter is fs, in the band that does not satisfy the sampling theorem, fs=(2/N)・fin...(6) (N=1 , 2, 3, 4,...), the measured value will oscillate as shown in FIG. However, even in a band that does not satisfy the sampling theorem, a normal measurement value can be obtained if the above-mentioned equation (6) is not satisfied.

For example, as the sample clocks CLK(1) to CLK(5) of each system in FIG. 1, as shown in FIG.
K(1)=(2/1.2)・finCLK(2)=(2
/1.1)・finCLK(3)=(2/1)・finCLK(4)=(2/0.9)・finCLK(5)=
(2/0.8)・fin is prepared, and the measured values obtained from each sample clock CLK(1) to CLK(5) are D(1), D(2), D(3), D(4),D
(5), D(3) oscillates between 0≦D(3)≦1. In addition, other data D(1), D(2), D
(4) and D(5) are values near the true value (=1/(2)1/2). Therefore, it can be seen that the true value can be obtained by removing the maximum value and minimum value from these D(1) to D(5) and averaging the remaining data. Note that the maximum and minimum values are removed when D(3) is 0≦D(3)≦21/
This is because it oscillates between 2.

[0035] These relationships will be explained using a concrete example.

CLK(1)=20.0KHzCLK(2
) = 20.5KHz CLK (3) = 21.0KHz CLK (4) = 19.5KHz CLK (5) = 19.0KHz is used to measure the 31.5KHz signal under measurement that is included in the band that does not normally satisfy the sampling theorem. Measure.

[0037] Here, the clock CLK (3) is as described above (
6) Since the condition of N=3 in the equation is satisfied, the measured value becomes indeterminate as 0≦measured value≦21/2.

However, the measured values using other sample clocks obtain true values (=1/(2)1/2). Therefore, as described above, by removing the maximum value and the minimum value and averaging the remaining data, the measurement data based on the clock CLK (3) system is removed and the true value of the effective value can be obtained.

FIG. 4 is a block diagram of another embodiment of the present invention, in which parts common to those in FIG. 1 are given the same reference numerals. The difference between FIG. 4 and FIG. 1 is that the A/D converter 2, square calculator 4, exponential average calculator 5, and square root calculator 6 are provided in only one system, and multiple sample clocks CLK(1) to CLK (n
) is applied to the A/D converter 2 by switching with a switch 9.

According to the configuration of FIG. 4, the measurement data D(1) based on each sample clock CLK(1) to CLK(n)
Although a time lag occurs in ~D(n), since there is only one measurement system, the cost is reduced compared to the configuration of FIG. 1.

[0041]

As described above in detail, according to the present invention, the following effects can be obtained. (1) By providing a plurality of sample clocks with different fixed periods to the A/D conversion means, normal effective value measurements can be made even for measurement signals in bands that do not satisfy the sampling theorem. (2) Since it is possible to use low-speed A/D converters, circuits can be easily designed without being subject to band constraints due to the sampling period limitations of conventional A/D converters, and are cost effective. can be lowered.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention.

FIG. 2 is a signal waveform diagram of each part in FIG. 1;

FIG. 3 is an explanatory diagram of the relationship between the sampling theorem and measured values.

FIG. 4 is a block diagram showing another embodiment of the present invention.

[Explanation of symbols]

1 Input terminal 2 A/D converter 3 Clock generator 4　　　　　　Square calculator 5 Exponential averaging calculator 6 Square root operator 7 Arithmetic unit (CPU) 8 Indicator 9 Switch

Claims (1)

[Claims] 1. A/D conversion means for sampling a signal under measurement at a plurality of different fixed periods; squaring calculation means for individually squaring sampling data at each fixed period of the A/D conversion means; Exponential average calculation means for individually performing an exponential average calculation on the square calculation data of the square calculation means, and square root calculation for individually performing a square root calculation on the exponential average data of the exponential average calculation means. 1. A broadband effective value measuring device comprising: a means for determining and removing a maximum value and a minimum value from square root data of the square root calculating means; and calculating means for calculating an average of the remaining data.