공기조화기에서 직렬아크방전 검출에 관한 연구

Abstract

This thesis dealt with the detection of series arc signal produced from an air conditioner’s power system. Series arc is a continuous, luminous electrical discharge across an insulation material. Since it progresses insulation breakdown when arc discharge lasts for a long time, it is important to prevent accidents due to electrical failures through the early detection of series arc. This paper, therefore, studies voltage, and current and optical detection methods.

Through the experiment, a module consisting of a filter with low cut-off frequency of 170kHz was applied to detect voltage signal, considering the frequency components of an arc signal. The current detection module consists of a high frequency current transformer (HFCT) with frequencies ranging from 20kHz to 30MHz and low-noise amplifier with the gain of 20dB and frequencies ranging from 60Hz to 47MHz at -3dB. An optical detection module consisting of an ultra violet(UV) sensor with measurement ranging from 210nm to 395nm and a low-noise amplifier with the gain of 40dB and frequencies ranging from 60Hz to 30MHz at -3dB was fabricated. Also, an algorithm was proposed to discriminate the arc signal only in the presence of noise, comparing with the number of pulses inputted per 100 ms.

The application experiment was carried out using an arc generator specified in UL1699, an impulse noise simulator and a lightning surge generator. From the results, the fabricated module and algorithm could detect the arc signals over 99% without any malfunctions against impulse noise of 2kV with a pulse width of 1000ns and surge current of 4 kA with a rise time of 8/20 us.