Projects / Programmes
New generation of industrial noise thermometers
| Code |
Science |
Field |
Subfield |
| 2.15.04 |
Engineering sciences and technologies |
Metrology |
Metrologies in areas |
| Code |
Science |
Field |
| P180 |
Natural sciences and mathematics |
Metrology, physical instrumentation |
| Code |
Science |
Field |
| 2.02 |
Engineering and Technology |
Electrical engineering, Electronic engineering, Information engineering |
noise thermometer, temperature, primary reference measurement procedure, uncertainty
Researchers (7)
Organisations (1)
Abstract
Noise thermometer is a primary thermometer, which means that it uses a primary measurement method – the temperature is calculated directly from the physical law that describes the thermal noise of an electrical resistor. Thermal or Johnson noise is caused by thermal motion of electrons on an electrical resistor and depends on temperature, electrical resistance and frequency bandwidth. As a primary thermometer, the noise thermometer was until now mainly used in a laboratory environment and at Boltzmann constant determinations with achievable uncertainties between 2·10-6 in 3·10-6. The main difficulties are especially the measurements of extremely low electrical voltages, the use of correlators and precise quantum voltage sources, the nonlinearity of amplifiers and very long measurement times. They can reach up to 100 days in order to minimize the statistical uncertainty. Such noise thermometer design is useless for day to day measurements of temperature where the measurement time should be much shorter.
The new noise thermometer design, useful for industrial applications, does not need a reference resistor or an input switch. It uses the pseudorandom voltage source of white noise – the sum of many sine waves of different frequencies with the same amplitude but random phase. This signal is stored in a digital form and is converted to analog form when needed. Such a noise thermometer design is not limited to small resistances and frequency bandwidths. Resistors with resistances of a few kΩ can be used and the frequency bandwidth can be as high as a few Mhz. Therefore, the RMS value of the thermal noise voltage is significantly higher. The measurement time can also be lowered to less than a minute.
The goal of the research project is to design and manufacture an industrial noise thermometer of a new generation with the use of the proposed pseudorandom noise source. This goal can be divided into the following stages. Firstly, the appropriate components must be selected, and the electrical circuit must be made. Secondly, the pseudorandom noise generation technique must be determined, with a sum of different combinations of sine waves. The last stage is the algorithm for correlator’s output signal processing in order to accurately measure the power of the Johnson noise on the measurement resistor. The target measurement uncertainty of the thermometer is 0.1 °C. Such uncertainty would enable a wide use in industry, where currently resistance thermometers and thermocouples are most commonly used. These thermometers are secondary thermometers and must be frequently calibrated (for example, annually) in order to track the changes in their behavior and their drift. On the other hand, a noise thermometer is a primary thermometer, which consequently means that is has to be calibrated at one temperature point only. Therefore, the time and cost of calibration is significantly lowered. The thermometer should not drift and should be immune to electromagnetic interference from the environment. The temperature range is from –50 °C to 300 °C, which is one of the most common temperature ranges where measurements in industry are made.
Significance for science
Successful design of the noise thermometer would open a possibility of wide primary thermometer use. The main advantage is the one time calibration in one temperature point. The measurements can then be performed in a wide temperature range without the change in uncertainty and without drift. A potential outcome of the project could be the use of virtually any electrical resistor as a thermometer in certain conditions.
Secondary (classical) thermometers that work on the principle of volume change, change in electrical resistance or change in the generated voltage, must be calibrated in the whole temperature range and the correction tables or suitable polynomial equations that completely describe the behavior of the thermometer must be determined. Due to various changes in thermometers, they have to be calibrated annually (or even more frequently if they are used in critical applications). Noise thermometer is a primary thermometer and needs to be calibrated in only one temperature point. The drift will be significantly lower and the calibration interval will be longer. Therefore, the cost of calibration of noise thermometers is almost negligible in comparison with the cost of secondary thermometer calibration.
Significance for the country
Successful design of the noise thermometer would open a possibility of wide primary thermometer use. The main advantage is the one time calibration in one temperature point. The measurements can then be performed in a wide temperature range without the change in uncertainty and without drift. A potential outcome of the project could be the use of virtually any electrical resistor as a thermometer in certain conditions.
Secondary (classical) thermometers that work on the principle of volume change, change in electrical resistance or change in the generated voltage, must be calibrated in the whole temperature range and the correction tables or suitable polynomial equations that completely describe the behavior of the thermometer must be determined. Due to various changes in thermometers, they have to be calibrated annually (or even more frequently if they are used in critical applications). Noise thermometer is a primary thermometer and needs to be calibrated in only one temperature point. The drift will be significantly lower and the calibration interval will be longer. Therefore, the cost of calibration of noise thermometers is almost negligible in comparison with the cost of secondary thermometer calibration.
