Noise Thermometer Research Articles

Systematic Errors in a High-Accuracy Johnson Noise Thermometer

The design of a cross-correlation noise thermometer based on a very-low-noise preamplifier and a high-speed, high-accuracy analog-to-digital converter is described. Error estimations indicate that the thermometer will have a basic accuracy better than 10 mK. Further self-check procedures are outlined that will reduce the systematic errors to approximately ±2 mK. The measurement time required to achieve a statistical uncertainty of ±3 mK is 2.5 days.A thermometer of the design suggested is under construction at the Physics and Engineering Laboratory of the DSIR, for operation between 0 °C and 200 °C.

Dedicated multiprocessor system for calculating Josephson-junction noise thermometer frequency variances at high speed

A Josephson-junction noise thermometer produces a sequence of frequency readings from whose variations the temperature of the thermometer may be calculated. A preprocessor system has been constructed to collect the frequency readings delivered to an IEEE 488 bus by an ordinary counter operating at up to 1000 readings per second, perform the required calculations, and send summary information to a desk calculator or minicomputer on another 488 bus at a more convenient rate.

Effect of additive noise and bandpass filter on the performance of a Josephson junction noise thermometer

A Josephson junction may be used to convert the voltage fluctuations generated by Johnson (i.e., thermal) noise in a resistor into frequency fluctuations. The variance of the frequency fluctuations is thus a measurement of the Johnson noise and, therefore, of the temperature. This particular type of noise thermometer has been used at the National Bureau of Standards as part of a program to define a cryogenic temperature scale. We have studied the detailed influence of two types of post-detection bandpass filters (square and one-pole Butterworth) on the measured variance. We report here on these measurements and the circuit model used to fit them. The conclusion is that small but well understood corrections must be applied to the measured variance whenever the measurement time (i.e., gate time) of the frequency becomes comparable with the response time (defined as the inverse of the bandwidth) of the bandpass filter.

Noise equivalent circuit of linear passive two-ports with applications to transmission lines

A new representation of the thermal noise of linear passive two-ports, expressed in terms of the Thevenin equivalent circuit and convenient to circuit designers, has been derived. This representation is applied to the determination of the contribution of the noise of a transmission line (a special case of a two-port) to the total noise of a noise thermometer system. Preliminary experimental results from measurement of temperature with a noise thermometer connected via a long transmission line encourages further development of practical noise thermometers.

Noise thermometer for nuclear reactors

Noise thermometer for nuclear reactors

Noise thermometer with inherent-noise compensation

Noise thermometer with inherent-noise compensation

Use of noise and quadrupole nuclear thermometers in thermometry

Use of noise and quadrupole nuclear thermometers in thermometry

Selectivity in measuring the thermodynamic temperature of an ionized gas with a noise thermometer based on a capacity transducer

Selectivity in measuring the thermodynamic temperature of an ionized gas with a noise thermometer based on a capacity transducer

Mathematical modelling of the impedance of a Josephson junction noise thermometer

Recent experimental work on noise thermometers consisting of a resistively shunted superconducting loop containing a Josephson junction (a resistive SQUID), has shown some novel behavior of the SQUID dc impedance as a function of rf power. We present a mathematical analysis of the intrinsic behavior of a resistive SQUID in the limit of negligible noise and negligible feedback to the rf circuit. A nonlinear, first-order differential equation is thought to be a reasonable descriptor of this system. Because the radio frequency driving the SQUID is much larger than the Josephson frequency, we are able to obtain a pair of equations in which no rf oscillating terms appear, and which are amenable to numerical solution. The dc impedance calculated from these equations has several, but not all, of the experimentally observed features.

Modeling the impedance of a Josephson-junction noise thermometer

Recent experiments on noise thermometers made with Josephson junctions [rf SQUID's (superconducting quantum interference devices) shunted by a small resistance $R$] have shown that the SQUID impedance has an envelope which generally rises to the value of $R$ as the rf power is increased, although the envelope can be distorted when the frequency is just above the rf circuit resonance. We summarize the results of a mathematical analysis which shows that in the absence of noise, and with no reactive feedback of the SQUID upon the rf circuit, the SQUID impedance will have a uniformly rising envelope because of the SQUID inductance. The calculated curves have the characteristics of the experimental observations in the rising-envelope cases.

A self-calibrating rhodium-iron resistive SQUID thermometer for the range below 0.5 K

We report on experiments with a prototype resistive SQUID device which show that it can serve both as a primary noise thermometer and as a secondary resistance thermometer in the range 0.01--0.52 K. The resistor in the circuit was made from an alloy of Rh with 0.5% Fe whose resistivity has an appreciable temperature dependence in this range. The high sensitivity of the SQUID allowed the resistance to be measured very accurately with negligible dissipation of heat. Since values of absolute temperature could be obtained by noise thermometry, the device was in effect a self-califrating resistance thermometer. This combination of features is a rarity in thermometry in general, and may be unique in this temperature range. A version of this new thermometer has been fabricated and tested in the range 0.01--0.52 K. The results of experiments with this prototype are described, its limitations are examined, and ways of improving it are outlined.

The establishment of a temperature scale from 0.01 K to 0.05 K using noise and 60Co γ-ray anisotropy thermometers

We report here on our latest intercomparison of the Josephson junction noise thermometer and 60Co γ-ray anisotropy thermometer from 0.01 K to 0.05 K. These results show both thermometers are in agreement to better than 0.5%. This is an improvement over our previous data and is due mainly to better temperature control and advances in instrumentation.

Tin, a candidate for low temperature NMR thermometer?

The nuclear spin-lattice relaxation times for the tin isotopes Sn117 and Sn119 have been measured in the temperature range between 20 and 50 mK with a SQUID magnetometer by watching the longitudinal magnetization of the nuclear spins after the application of a rf pulse. The temperature was determined using a SQUID noise thermometer. The Korringa constants were estimated from the τ1 values. The experimental device and the principal methods are described and the results and thermometer applications discussed.

Measurement of the Thermodynamic Temperature of 4He at Various Vapour Pressures by a Noise Thermometer

A noise thermometer has been developed, which operates by the correlation method suggested by van der Ziel. The temperature of the liquid 4He at seven different vapour pressures was measured with an accuracy of about 10-4. Within 0.4 mK the noise temperatures coincide with the new T76-scale

Absolute high-temperature Johnson noise thermometry

A switching noise thermometer incorporating a series-resonant filter between the sensor and reference resistors and the amplifier stages is described. The first-order dependence of the temperature measurement on the magnitudes of the sensor and reference resistances and also on the magnitude of any stray capacitance is thereby removed. The DC voltage developed across a shot-noise diode load (the reference resistor) gives a direct measurement of the difference between the absolute temperatures of the sensor and reference resistors. The configuration enables noise signal bandwidths of 0.5 MHz to be used, allowing temperatures to be measured with a statistical uncertainty of 0.2% with a 1 s measurement time. FETS are incorporated at the input of the system in place of the conventional electromechanical switches.

Low frequency impedance and noise properties of an RF biased resistive SQUID

A noise thermometer consisting of a nonhysteretic resistive SQUID has been used for unusually precise measurements of the low frequency Josephson linewidth of a heavily shunted point-contact. The results agree with theories appropriate for weak links, and support the contention that the pair fluctuation noise calculated for tunnel junctions is not present in point-contacts.

Josephson-effect absolute noise thermometer: Resolution of unmodeled errors

An absolute cryogenic temperature scale is being defined at the National Bureau of Standards using the linewidth of radiation emitted by a resistively shunted Josephson junction. We report careful measurements of the impedance of this noise thermometer and show that its behavior is in excellent agreement with the predictions of the resistively shunted junction model assuming a sinusoidal current-phase relationship. The implications for absolute noise thermometry are discussed.

Noise Thermometry in the Range 630-962°C

A noise thermometer has been developed at IMGC and has been used to realize the Thermodynamic Scale (TS). Differences between TS, as realized at IMGC, and the IPTS-68 were determined in the range 630-962°C and the results are presented.The discussion of results primarily deals with the factors introducing measurement uncertainties and with the comparison of noise thermometer results with those of recent radiation thermometry and gas thermometry experiments.Data fitting provides the TS versus IPTS-68 differences as a continuous function of the Celsius temperature in the range 630-1064°C. For this, data from monochromatic pyrometry experiments are taken into account in order to cover the range from the freezing point of silver to the freezing point of gold.

Undergraduate experiment on noise thermometry

An absolute temperature scale noise thermometer is described and Boltzmann’s constant is measured. The apparatus involves an easily constructed temperature probe and standard electronic instruments. The design considerations for the experiment are a useful introduction to low noise electronics.

Thermometry in 3He

Some of the methods used in measuring the temperatures of the transitions in 3He are reviewed. Although noise thermometers, γ-ray anisotropy thermometers and even Pt NMR thermometers are satisfactory for furnishing temperature standards, they are not satisfactory secondary thermometers. The susceptibility of CMN and various 3He properties make more precise and quicker thermometric indications.