JPH03209177A - Measurement of insulation resistance - Google Patents

Abstract

PURPOSE:To separate and remove noise components adjacent to a signal frequency for measurement and to attain the measurement with high accuracy by subjecting to a high speed Fourier transformation the waveform data converted to digital values from the output of filter at a constant time interval. CONSTITUTION:A leak current passing through an insulation resistance Rg and an electrostatic capacitance Cg to the earth is generated by a low frequency signal for measurement impressed to the earth line through an injection transformer. This leak current is detected by a current transformer and inputted to an A/D conversion part through the filter. A control signal for A/D conversion is sent out to the A/D conversion part by a control part with making the low frequency signal as a trigger. The digital data subjected to A/D conversion are stored in a specified address by a waveform storing part. This address is decided by the control part. At the time point when a series of A/D conversion is finished and a storage of digital data into the waveform storing part is finished, the high speed Fourier transformation is performed for the digital data in the waveform storing part by an arithmetic part in accordance with a calculation executing signal of the control part. Further, real number parts only are extracted among the frequency components same as those of low frequency signal for measurement, then outputted.

Description

[Detailed Description of the Invention] [Field of Industrial Application] The present invention relates to a method for measuring the insulation resistance of an electrical circuit under live conditions, and in particular to a method for measuring the transient response and waves of current transformers, which cannot be ignored when measuring large insulation resistance values. This paper develops an insulation resistance measurement method that compensates for the effects of noise frequency components near the measurement low-frequency signal frequency that cannot be removed by any means. [Prior Art] A conventional method for measuring insulation resistance is shown in FIG. 2(a).

The signal generated by the measurement low frequency signal generator is injected into the grounding wire via the injection transformer, and this signal voltage changes the current that flows through the insulation resistance R0 and ground capacitance C, and returns to the grounding wire. A method is used to measure the insulation resistance of an electrical circuit by detecting it with a flowmeter, extracting only the leakage current component of the measurement low-frequency signal with a filter, and then detecting only the insulation resistance component with a synchronous rectifier. In addition, the synchronization signal for synchronous rectification needs to be in phase with the measurement low-frequency signal voltage injected into the grounding wire via the injection transformer, and the leakage current of the measurement low-frequency signal is caused by the current transformer and , the phase is shifted from the measurement low frequency signal voltage injected by the filter, so a common method is to provide a phase correction section to correct the shifted phase difference.

[Problem that the invention seeks to solve]

Traditional! ! The measurable insulation resistance value when using the edge resistance measurement method is determined by the human power equivalent noise voltage of the filter following the current transformer and the magnitude of the noise voltage included within the passband width.

First, Figure 2 shows an equivalent circuit of the electrical circuit including the insulation resistance Rg, the injection transformer, and the current transformer regarding the human power equivalent noise voltage (
This will be explained using b).

The voltage injected into the electrical circuit by the injection transformer is generally 0.
When the insulation resistance R is about 5V, the leakage current I1 is 5μA (0.5V/100kΩ), and the number of turns of the secondary winding of a current transformer is generally about 2000 turns, so current transformation The secondary current I2 of the device is approximately 2.5nA (5μA/
2000 turns).

Therefore, the voltage across the current transformer terminating resistor RL is Rt=100
Approximately 250nV as Ω (=100Ω×2.5r+
A). Increasing the termination resistor RL can increase the voltage across RL, but the current transformer's 1
Generally, R
, is used below 100Ω.

In addition, the input equivalent noise of commercially available ultra-low noise operational amplifiers is 3 to 5 nV at best, which is 1/1 of the signal component.
This means that there is a noise component of 0.00 to 1/50. High precision,
When performing high stability and high insulation resistance detection, the noise component must be less than 1/1000 of the signal component.
With conventional methods, it is impossible to detect insulation resistance Rg>100kΩ with high precision and high stability. Next, the noise voltage included within the passband width of the filter will be explained.

Generally, a bandpass filter is used to pass the measurement low-frequency signal, and the passband is set to a constant frequency width centered around the measurement low-frequency signal frequency. Since the passband of this filter contains a noise component due to the input converted noise voltage, the passband width of the filter may be narrowed in order to reduce the noise component in the output of the filter.

However, when the passband width is narrowed, the cutoff frequency of the filter becomes close to the measurement low frequency signal frequency, and the input voltage/output voltage phase characteristic of the filter with respect to temperature of the measurement low frequency signal frequency deteriorates.

As a result, the passband width is unnecessarily narrowed, and a noise component due to the input equivalent noise voltage is generated in the output of the filter.When detecting high-voltage gauze resistance, it is necessary to separate the low-frequency signal for measurement and the noise component. There was a drawback that it was difficult to do so.

[Means to solve the problem]

Means for solving the problem will be explained using FIG. 1(a).

The leakage current component of the electrical circuit detected by the current transformer is input to the analog/digital converter (hereinafter referred to as the A/D converter) via the filter. Further, the A/D conversion section is controlled by the control section, quantizes the analog input using the measurement low frequency signal as a trigger signal, and then stores the quantized waveform data in the waveform storage section. This relationship is shown in FIG. 1(b).

The control section generates an A/D conversion control signal for a time T2 at intervals of time T1 using the zero voltage point A of the measurement low frequency signal as a trigger point, and applies it to the A/D conversion section. The A/D converter converts the analog input signal into a digital value each time an A/D conversion control signal is applied, and sends the digital value to the waveform storage. The waveform storage section is
A in the address determined by the memory address setting signal set in advance for each A/D conversion control signal generated from the control unit.
/Stores the digital value of the D converter output. That is, analog/digital conversion is performed (T2/T++1) times during time T2, and waveform digital data is stored in the storage locations prepared for this number of times. Furthermore, the number of times is preferably 2' (n is a natural number) in view of the fast Fourier transform to be performed in the subsequent calculation unit, but there is no need to specify this in particular. Furthermore, for the sake of explanation, the trigger point was set to zero voltage point A, but in order to compensate for the phase delay of the analog signal due to the current transformer and filter, the trigger point was moved by the phase delay and set to a point other than the zero voltage point. I don't mind. The digital waveform data stored in the waveform storage section is subjected to fast Fourier transform by the arithmetic section. Fast Fourier transform is a general method for converting a series of real data in the time domain into data in the frequency domain.

In other words, by performing fast Fourier transform on waveform digital data, it is possible to obtain calculation results by dividing the signal component of a specific frequency into a real part and an imaginary part. Furthermore, a low frequency signal frequency is selected as this specific frequency, and in the calculation result, the real part becomes a component with the same phase as the measurement low frequency signal used as a trigger signal, and the imaginary part has a π/2 phase with the same low frequency signal. Since these are different components, it is possible to distinguish whether the leakage current component is due to the insulation resistance R or the leakage current component due to the ground capacitance c9. In other words, when the trigger voltage point is zero voltage, the real part after fast Fourier transform becomes the leakage current component due to the insulation resistance R9, and the imaginary part becomes the ground capacitance C. This becomes a leakage current component.

However, this is limited to cases where there is no phase delay due to current transformers or filters.

If there is an in-phase delay, a circuit for correcting the phase delay may be provided after the filter, or the trigger voltage point may be changed. Generally, the frequency resolution when performing fast Fourier transform is the A/D conversion control signal T2 time and the number of A/D conversions N.
It depends on ("1+T2/T) and is given by (2/(T2XN)).

Fast Fourier transform is a technique for detecting specific frequency components, but expressed in another way, it is a bandpass filter that passes only specific frequency components, and the passband width is determined by the frequency resolution, that is, the A/D conversion control signal T2. Time and number of A/D conversions N
Determined by

As described above, the wave technology using fast Fourier transform is a digital filter, and it is possible to easily realize a filter with a narrow passband width with high precision and high stability without depending on environmental changes such as temperature and humidity. Furthermore, the phase characteristics with respect to temperature changes can be ignored, and the analog filter before the A/D converter can be further simplified, making it possible to provide a device with high stability against environmental changes.

〔Example〕

In Fig. 1(a), a leakage current is generated through the insulation resistance R and the ground capacitance C9 due to the measurement low frequency signal applied to the ground wire through the injection transformer. This leakage current is detected by a current transformer and is manually input to the A/D converter through a filter. Further, the control section uses the low frequency signal as a trigger to send out an A/D conversion control signal to the A/D conversion section as described above. The waveform storage unit stores A/D converted digital data at a predetermined address. This address is determined by the control section. series of A's
When the /D conversion is completed and the storage of digital data in the waveform storage unit is completed, the control unit sends a calculation execution signal to the calculation unit, and the calculation unit converts the digital data in the waveform storage unit at high speed. Fourier transform. Further, from the results of the fast Fourier transform, only the real part is extracted from the frequency components that are the same as the measurement low frequency signal frequency and output. This output is inversely proportional to the value of the insulation resistance R9 of the electrical circuit. In addition, numerical processing such as calculating the average value of the number of outputs can be easily carried out in order to achieve higher accuracy.

〔Effect of the invention〕

■ Conventional same y. Compared to the uu-style method, the present invention, which performs digital calculations after AID conversion, allows l! I! ! The number of parts is greatly reduced and assembly work efficiency can be improved.

■ By using the fast Fourier transform method, it is possible to separate and remove noise components near the measurement signal frequency that could not be removed by a filter. ■ By using fast Fourier transform and digital processing, it is possible to measure high insulation resistance with high stability and high accuracy without being affected by changes in phase characteristics due to environmental changes.

[Brief explanation of drawings]

Figure 1 (a) is a diagram showing the concept and embodiments of the present invention, Figure 1 (b) is a diagram showing timing for explaining the concept, Figure 2 (a) is an insulation resistance measuring device using a conventional method. FIG. 2(b) is an equivalent circuit diagram for explaining the problems of the conventional measurement method.

Claims (1)

[Claims] A low-frequency signal voltage for measurement different from the commercial frequency is applied to the grounding wire of the transformer via an injection transformer to the power line, and a leakage current component of the low-frequency signal for measurement is detected, and at the same time, the detected unnecessary current component is After removal by a filter, the output of the filter is converted into a digital value at a fixed period and at fixed time intervals using a synchronization signal obtained by waveform shaping from the measurement low frequency signal as a trigger and stored in the waveform data storage unit. An insulation resistance measuring method characterized in that an insulation resistance component is separated and detected from a measurement signal frequency component by performing fast Fourier transform in a calculation unit using this waveform data.