A high-precision temperature measurement system based on noise cancellation

Weak signal detection is a major problem in the signal processing. The paper proposes a method, in which the combination of hardware compensation and the corresponding algorithm cancellation are used to overcome the noise drifts and improve the detection accuracy of the high-precision temperature measurement system. A high-precision temperature measurement system is designed, platinum resistor PT1000, used as the temperature sensor, is in series with reference resistor and driven by constant-current source, which efficiently eliminates the effects of lead resistance and self-heating. A Discrete Gaussian noise model is established based on this hardware system. After the sampling signal is amplified, the A/D conversion is taken, then the temperature drifts correction is taken through SCM software and the noise cancellation algorithm is used by the host computer, solving the problems of drifts and noise and achieving better precision and stability. The experimental results show that this method greatly improves the accuracy and stability of the temperature measurement system and the measurement error is less than 0.007°C. Moreover, the standard deviation is less than 0.005°C.