Analysis of an R-SQUID noise thermometer:Dissertation

In this thesis the metrological performances of the R SQUID intended to be used as a cryogenic thermometer is verified.The R SQUID consists of the Josephson junction which converts voltage noise from the shunt resistor into frequency fluctuations.Measuring the value of the shunt resistor and the variance of the frequency fluctuations, the temperature of the bath can be determined.Unfortunately, the absolute temperature determination is disturbed by several effects: a rolloff introduced by a post detection filter, additional noise caused by the rf amplifiers, a mixed noise effect caused by the nonlinearity of the Josephson junction together with rf noise in the tank circuit, and the limited number of the frequency counts.In addition, the determination of the shunt resistor may cause a systematic error.In this work the influence of the post detection filter and the additive noise on the frequency variance is analyzed and the analytic models for the dc impedance and the noise performances of the R SQUID are derived.All our theoretical models are nowadays experimentally verified; some of the experiments are reported in this work.The analytic models are used to estimate the total inaccuracy of the noise thermometer and to find the optimal choice of the device parameters.The work reveals that the ultimate accuracy of the temperature determination is proportional to the total measurement time and the pump frequency and inversely proportional to the noise temperature of the preamplifier.This work shows that the R SQUID is complex gauge to maintain a cryogenic temperature scale.However, the careful choice of the device parameter and the point of operation, the use of the low noise cooled preamplifier, and the operation at high pump frequencies, enable very accurate determination of the temperature of the bath in which the R SQUID is immersed