Instrumentation Sensors Book

of ±1°C or better. Other advantages are that electronics can be integrated onto the same die as the sensor, giving high sensitivity, easy interfacing to control systems, and making possible different digital output configurations [10]. The thermal time constant varies from 1 to 5 seconds, and internal dissipation can cause up to a 0.5°C offset. Semiconductor devices also are rugged, have good longevity, and are inexpensive. For the above reasons, the semiconductor sensor is used extensively in many applications, including the replacement of the mercury in glass thermometer.

10.4Application Considerations

Many devices are available for temperature measurement. The selection of a device will be determined by the needs of the application. A comparison of the characteristics of temperature-measuring devices is given. Thermal time constant considerations, installation calibration, and protection also are discussed.

10.4.1Selection

In process control, a wide selection of temperature sensors is available [11]. However, the required range, linearity, and accuracy can limit the selection. In the final selection of a sensor, other factors may have to be taken into consideration, such as remote indication, error correction, calibration, vibration sensitivity, size, response time, longevity, maintenance requirements, and cost [12]. The choice of sensor devices in instrumentation should not be degraded from a cost standpoint. Process control is only as good as the monitoring elements.

10.4.2Range and Accuracy

Table 10.8 gives the temperature ranges and accuracies of temperature sensors, the accuracies shown are with minimal error correction. The ranges in some cases can be extended with the use of new materials. Table 10.9 gives a summary of temperature sensor characteristics.

Table 10.8 Temperature Range and Accuracy of Temperature Sensors

10.4.3Thermal Time Constant

A temperature detector does not react immediately to a change in temperature. The reaction time of the sensor, or thermal time constant, is a measure of the time it takes for the sensor to stabilize internally to the external temperature change, and is determined by the thermal mass and thermal conduction resistance of the device [13]. Thermometer bulb size, probe size, or protection well can affect the response time of the reading. For example, a large bulb contains more liquid for better sensitivity, but this will increase the time constant to fully respond to a temperature change.

The thermal time constant is related to the thermal parameters by the following equation:

where tc is the thermal time constant, m is the mass, c is the specific heat, k is the heat transfer coefficient, and A is the area of thermal contact.

When temperature changes rapidly, the temperature output reading of a thermal sensor is given by:

where T is the temperature reading, T1 is the initial temperature, T2 is the true system temperature, and t is the time from when the change occurred.

The time constant of a system tc is defined as the time it takes for the system to reach 63.2% of its final temperature value after a temperature change. As an example, a copper block is held in an ice-water bath until its temperature has stabilized at 0°C. It is then removed and placed in a 100°C steam bath. The temperature of the copper block will not immediately go to 100°C, but its temperature will rise on an exponential curve as it absorbs energy from the steam, until after some time period (i.e., its time constant), it will reach 63.2°C, eventually reaching 100°C. During the second time constant, the copper will rise another 63.2% of the remaining temperature to get to equilibrium. That is, (100 − 63.2) × 63.2% = 23.3°C, or at the end of two time constant periods, the temperature of the copper will be 86.5°C. At the end of three periods, the temperature will be 95°C, and so on. Where a fast response

time is required, thermal time constants can be a serious problem. In some cases, the constants can be several seconds in duration, and a correction may have to be electronically applied to the output reading to obtain a faster response. Measuring the rate of rise of the temperature indicated by the sensor and extrapolating the actual aiming temperature can do this. The thermal time constant of a body is similar to an electrical time constant, which is discussed in the chapter on electricity under electrical time constants.

10.4.4Installation

Care must be taken in locating the sensing portion of the temperature sensor. It should be fully encompassed by the medium whose temperature is being measured, and not be in contact with the walls of the container. The sensor should be screened from reflected and radiant heat, if necessary. The sensor also should be placed downstream from fluids being mixed, to ensure that the temperature has stabilized, but as close as possible to the point of mixing, to give as fast as possible temperature measurement for good control. A low thermal time constant in the sensor is necessary for a quick response [14].

Compensation and calibration may be necessary when using pressure-spring devices with long tubes, especially when accurate readings are required.

10.4.5Calibration

Temperature calibration can be performed on most temperature sensing devices by immersing them in known temperature standards, which are the equilibrium points of solid/liquid or liquid/gas mixtures (also known as the triple point). Some of these standards are given in Table 10.10. Most temperature-sensing devices are rugged and reliable, but they can go out of calibration, due to leakage during use or contamination during manufacture, and therefore should be checked on a regular basis.

10.4.6Protection

In some applications, temperature-sensing devices are placed in thermowells or enclosures, to prevent mechanical damage or for ease of replacement. This kind of protection can greatly increase the system response time, which in some circumstances may be unacceptable. Sensors also may need to be protected from overtemperature, so that a second, more rugged device may be needed to protect the main sensing device. Semiconductor devices may have built-in overtemperature protection. A fail-safe mechanism also may be incorporated for system shutdown, when processing volatile or corrosive materials.

Table 10.10 Temperature Scale Calibration Points

10.5Summary

Temperature is the most important physical parameter, because all other physical parameters are temperature-dependent. Temperature can be measured using Celsius or Kelvin, in the SI system of units, and Fahrenheit or Rankine in the English system of units. This chapter described the relations between the units and introduced the basic concept that temperature is a measure of the heat energy contained in a body or the amplitude of molecular vibration. The vibration amplitude and molecular attraction determines the phase of the material. Heat energy can be measured in British thermal units or joules, and can be transferred between bodies either by conduction, convection, or radiation. The mechanics of heat transfer was described.

A large number of instruments is available for temperature measurement. The choice of instrument is determined by the requirements of the application. New innovations, such as the semiconductor digital thermometer, have brought about the demise of the mercury thermometer. Inexpensive On/Off applications used bimetallic devices, but these are also being replaced by semiconductor devices. Wide temperature range applications typically use thermocouples, and high accuracy devices are RTDs. Some devices have long settling times due to their thermal time constant, which can be compensated for electronically [15].

Definitions

Absolute zero is the temperature at which all molecular motion ceases, or the energy of a molecule is zero.

British thermal unit (Btu) is defined as the amount of energy required to raise the temperature of 1 lb of pure water 1°F, at 68°F and 1 atm.

Calorie unit (SI) is defined as the amount of energy required to raise the temperature of 1g of pure water 1°C, at 4°C at 1 atm.

Celsius or Centigrade scale (°C) uses 0° and 100° (100° range) for the freezing and boiling points, respectively, of pure water at 1 atm. Fahrenheit scale (°F) uses 32° and 212° (180° range) as the freezing and boiling points, respectively, of pure water at 1 atm.

Heat is a form of energy, and is a measure of the vibration amplitude of its molecules, which is indicated by its temperature.

Joules (SI) are units of heat energy.

Kelvin scale (K) is referenced to absolute zero, and based on the Celsius scale.

Rankine scale (°R) is a temperature scale referenced to absolute zero, and based on the Fahrenheit scale.

Sublimation is the direct transition from the gas state to the solid state without going into the liquid state, or the direct transition from the solid state to the gas state.

References

[1]Mathews, D., “Choosing and Using a Temperature Sensor,” Sensors Magazine, Vol. 17, No. 1, January 2000.

[2]Krysciar, T., “Low Temperature Measurements with Thin Film Platinum Resistance Elements,” Sensors Magazine, Vol. 16, No. 1, January 1999.

[3]Lavenuth, G., “Negative Temperature Coefficient Thermistors,” Sensors Magazine, Vol. 14, No. 5, May 1997.

[4]Humphries, J. T., and L. P. Sheets, Industrial Electronics, 4th ed., Delmar, 1993, pp. 320–323.

[5]Klopfenstien, Jr., R., “Linearization of a Thermocouple,” Sensors Magazine, Vol. 14, No. 12, December 1997.

[6]Bedrossian, Jr., J., “Infrared Noncontact Temperature Measurement,” Proceedings Sensor Expo, 1994, pp. 597–602.

[7]Barron, W. R., “Infrared Thermometry—Today,” Proceedings Sensor Expo, 1993, pp. 1–12.

[8]DeWitt, D. P., “Infrared Thermometry—Tomorrow,” Proceedings Sensor Expo, 1993, pp. 13–16.

[9]Ristic, L., Sensor Technology and Devices, Artech House 1994, pp. 287–315.

[10]Scolio, J., “Temperature Sensor IC’s Simplify Designs,” Sensors Magazine, Vol. 17, No. 1, January 2000.

[11]Maxwell, D., and R. Williamson, “Wireless Temperature Monitoring in Remote Systems,” Sensors Magazine, Vol. 19, No. 10, October 2002.

[12]Desmarais, R., and J. Breuer, “How to Select the Right Temperature Sensor,” Sensors Magazine, Vol. 18, No. 1, January 2001.

[13]Johnson, C. D., Process Control Instrumentation Technology, 7th ed., Prentice Hall, 2003, pp. 36–38.

[14]Paluch, R., “Field Installation of Thermocouple & RTD Temperature Sensor Assemblies,” Sensors Magazine, Vol. 19, No. 8, August 2002.

[15]Stokes J., and G. Palmer, “A Fiber Optic Temperature Sensor,” Sensors Magazine, Vol. 19, No. 8, August 2002.

where K is the dielectric of the material between the plates, A is the area of the plates in square meters, and d is the distance between the plates in meters.

There are three methods of changing the capacitance, as shown in Figure 11.3:

(1) changing the distance between the plates, (2) moving one plate with respect to the other plate to reduce the overlapping area between the plates, and (3) moving a dielectric material between fixed plates. Typically, differential sensing is used, as shown in Figure 7.5(a). Capacitance variation is a very accurate method of measuring displacement, and is used extensively in micromechanical devices where the distance is small, giving high capacitance per unit area. Capacitive devices are used in the measurement of pressure, acceleration, and level.

Light interference lasers are used for very accurate incremental position measurements. Monochromatic (i.e., single frequency) light can be generated with a laser and collimated into a narrow beam. The light from the laser beam is split and a percentage goes to a detector, but the main beam goes to a mirror attached to an object whose change in distance is being measured, as shown in Figure 11.4(a). The reflected beam is then directed to the detector via the beamsplitter and a mirror. A change in the position of the object of one-quarter wavelength increases both the incident and reflected beam length one-quarter wavelength, giving a change at the detector of one-half wavelength. When the reflected beam is in phase with the incident beam, (d) is N times an even number of quarter-wavelengths of the laser beam, the light amplitudes add, and an output is obtained from the detector. When the distance (d) is N times an odd number of quarter-wavelengths, the beam amplitudes subtract, and the output from the detector is zero, as shown in Figure 11.4(b). The movement of the object generates interference fringes between the incident light and the reflected light. These fringes can be counted to give the distance the object moves. The wavelength of the light generated by a laser is about 5 × 10−7m, so that relative positioning to one-quarter wavelength (0.125m) over a distance of 50 cm to 1m is achievable.

Ultrasonic, Infrared, Laser, and Microwave Devices can be used for distance measurement. The time for a pulse of energy to travel to an object and be reflected back to a receiver is measured, from which the distance can be calculated. The speed

Figure 11.3 Variable capacitors, with varying (a) distance between plates, (b) plate overlap, and

(c) dielectric.