Design Of Measurement System For Dissolved Gases Content In Transformer Oil

The amount of characteristic gases dissolved in transformer oil implies many information about the operating states of the transformer. The quantitative analysis of the characteristic gases can predict the fault condition of the transformer in time, so as to avoid the occurrence of more serious accidents. The photoacoustic spectrometry(PAS) gas-in-oil detecting method overcomes many flaws of the traditional methods, and it has such advantages as good selectivity, high sensibility and large dynamic range. In this thesis, a PAS trace gas detecting system is designed and which realized the accurately measurement of the trace gas content in the transformer oil.Based on the working principle of the photoacoustic effect, the infrared spectrum absorption characterization of the gases, and combining the characteristics of the infrared light, the micro-acoustic sensor and the photoacoustic cell with the typical characteristics of the photoacoustic signal, a PAS trace gas detecting system, which can control the modulation frequency of the infrared source through electric pulse, is designed. The system significantly reduced the complexity and the volume of the system, and it is suitable for the industrial field.To solve the problem that the photoacoustic signal of the non-resonant gas-in-oil is extremely weak, a high-precision Sigma-Delta ADC lock-in amplifying circuit is designed to detect the trace gas. This numerical lock-in amplifier is based on FPGA(Field Programmable Gate Array), and this amplifier is of low energy consumption and high reliability. The test results show that the detecting circuit can detect the microvolt signal effectively.Applying this PAS trace gas dissolved trace gas in transformer oil detecting system, different concentration of C2H2 is detected. To analyze the amplitude of the acquired photoacoustic signal, the detection error is within the range permitted. Meanwhile, the all results indicate that the PAS detecting system can detect the dissolved trace gas in oil effectively, can provide quantitative analysis of the trace gas, and can help to predict the possible fault of the transformer.