Industrial Temperature Primer

Chapter 1.

Temperature Measurement Devices:

Thermometers -

The first thermometer was invented by Galileo (1564-1642). It was an air thermometer consisting of a glass bulb with a long tube attached. The tube was dipped into a cooled liquid, then the bulb was warmed, expanding the air inside. As the air continued to expand, some of it escaped. When the heat was removed, the remaining air contracted causing the liquid to rise in the tube indicating a change in temperature. This type of thermometer is quite sensitive, but not too practical since it can be affected by the slightest change in atmospheric pressure.

In 1714 a gentleman by the name of Gabriel D. Fahrenheit invented both the mercury and the alcohol thermometer with which we are all quite familiar. Fahrenheit's mercury thermometer consists of a capillary tube which after being filled with mercury is heated to expand the mercury and expel the air from the tube. The tube is then sealed, leaving the mercury free to expand and contract with temperature changes. Although the mercury thermometer is not as sensitive as the air thermometer, it is not affected by the atmospheric pressure changes.

The mercury thermometer has a serious drawback, however. Mercury freezes at -39° Celsius, so it cannot be used to measure temperature below this point. Alcohol, on the other hand, freezes at -113° Celsius. Therefore, by substituting alcohol for mercury, much lower temperatures may be measured.

Many industrial thermometers register temperature by means of a pointer on a calibrated dial. These thermometers contain no liquid but do operate on the principle of unequal expansion. Since different metals expand at different rates, we can bond one metal to another and see that when heated, the bonded metal will bend in one direction and when cooled it will bend in the opposite direction (therefore the term "Bimetallic Thermometer"). This bending motion is transmitted by a suitable mechanical linkage to a pointer that moves across a calibrated scale. Although not as accurate as liquid in glass thermometers, BiMets are much more rugged, easy to read, and have a wider span making them ideal for many industrial applications. (Figure 1)

The filled system type of thermometer works on the same basic principle as the bimetallic thermometer. The sensing element is a capillary tube filled with a liquid or gas which expands with an increase in temperature. This sensing element delivers a motion of physical change that is applied to the control element which either indicates, records, or by comparing the signal to a setpoint can be used to control the temperature of a process.

The filled system thermometer differs from other thermometer types in that it can be used in some simple control situations. (Figure 2)

Thermocouples - History and Theory of Operation

In the year 1821, a very important discovery in the field of Thermometry was made. T. J. Seebeck observed that if two dissimilar metals are joined together to form a closed loop, and if one junction is kept at a different temperature from the other, an electromotive force is generated (called the Seebeck emf in honor of its discoverer) and electric current will flow in the closed loop. Experiments by Seebeck and others have shown that the amount of electric current flowing in the loop is relative in a predictable manner to the difference in temperature between the two junctions. So, if the temperature of one junction is kept at a known value, the temperature of the other junction can be determined by the amount of voltage produced. This discovery resulted in the temperature sensor that we know as the thermocouple. (Figure 3)

There are a couple of important laws governing the operation of Thermocouples. First, the Law of Homogeneous Circuits states that if thermocouple conductors are homogenous, they are unaffected by intermediate temperatures. If a junction of two dissimilar metals is maintained at T1 while the other is at T2, the thermal emf developed is independent and unaffected by any temperature distribution along the wires T3 and T4. This law forms the basis for the use of thermocouple grade extension wire. (Figure 4)

In figure 5. we see that because of the Law of Homogeneous Circuits, if the thermocouple wire is homogeneous, and if junctions T3 and T4 is less than or greater than T1, no affect in the measurement output emf will be seen. (Figure 5)

The second important law in thermocouple temperature measurement is the Law of Intermediate Metals. The Law of Intermediate Metals states that a third metal can be introduced into the circuit without creating errors if the junctions of the third metal to the thermocouple conductors are at the same temperature. When using thermocouples, it is usually necessary somewhere in the loop to introduce additional metals into the circuit. This happens when an instrument is used to measure the output of the thermocouple and the instrument input terminals are of a different metal (usually brass or gold) and when the junction is brazed or welded. It would seem that this introduction of other metals would change the emf output of the thermocouple and add error to the signal. However, as long as the junction of the third metal with the other two metals are at the same temperature, no error signal is produced. (Figure 6)

The law of intermediate metals comes into play when, for example, one uses a thermocouple with a brass terminal block in the connection head. The wires from the thermocouple element are attached to the terminal block to connect to the wires to the instrument. As long as there is no temperature gradient across the terminal block, no error will be introduced and the emf in the circuit will remain unaffected.

In figure 6 we see two dissimilar metals A and B with their junctions at T1 and T2 and a third metal C. joined on one leg. If C is kept at a uniform temperature along its length, the total emf in the circuit will not be affected.

A good example of the practical use of the Law of Intermediate Metals is shown is figure 7. In industrial applications, thermocouples are often terminated in a junction box where they are joined to thermocouple extension wire which extends back to the instrumentation in the control room. The terminal blocks used in the junction box are often constructed of a metal such as brass or nickel plated copper. Either of these metals represents an intermediate metal. According to the Law of Intermediate metals, as long as there is no difference in temperature between terminals T1 and T2 and between terminals T3 and T4, there will be no error introduced into the circuit by the intermediate metals. (Figure 7)

Thermocouple Types

Any two dissimilar metals can in theory be made into thermocouples. However, certain metals have been selected over time that make ideal thermocouples for various applications. These metals have been chosen for their emf output and their ability to operate under various conditions. There are several types of these "standard" thermocouples in use today.

As you can see from the chart on the next few pages, particular ISA calibrations of thermocouples are more suited to specific applications and temperature ranges than others.

One advantage thermocouples have over some other sensors is the ability to construct the sensor so as to suit most any application. From the simplest bare wire thermocouple, to protected sensors housed in protection tubes and thermowells with any number of different mounting arrangements. The fact that thermocouples are typically rugged, inexpensive, highly responsive and have a very broad temperature range makes them the temperature sensor of choice in many applications. Another important advantage is that since thermocouples measure temperature at the junction of the two dissimilar metals which are usually in the form of fine wire, they are naturally quite "tip sensitive". This means that you can measure temperature at a very small point of reference. One of the disadvantages of the thermocouple is that their outputs are quite non linear. Therefore, instruments used to measure temperature by thermocouples must include linearization circuitry. Also since the accuracy of the thermocouples is dependent upon the control of the alloy used in their manufacture, accuracy is somewhat limited.

(Click here to view a Standard ISA Thermocouple Calibration table.)

The Type "J" or iron-constantan thermocouple is the most widely used calibration of thermocouple. Over two hundred tons of iron and constantan materials are used in the manufacture of this calibration each year in the United States alone.

This popularity is in spite of the fact that the use of iron as a thermocouple material was vigorously opposed by many in the field of thermometry. Burgess and Le Chatelier, in their book on "Measurement of High Temperature" (1912), are emphatic on this point, basing their objections on the inhomogeneity of iron wires, and the consequent large parasitic emfs developed where a temperature gradient exists in the wire. However, the relatively high emf output, a comparatively low cost, and the adaptability to both oxidizing and reducing atmospheres justify iron-constantan's wide use. Also, the iron used in today's thermocouples are not appreciably less homogeneous than the constantan with which it is paired. When used under conditions where the temperature gradient along the wire is not subject to rapid fluctuations, the parasitic emfs seldom result in errors larger than one or two degrees F in measured temperature.

Constantan is an alloy of copper and nickel with a typical composition Cu57Ni43 plus the addition of small percentages of Mn and Fe. The precise composition of the alloy is not specifically defined and depends on whether it is to be used with iron for type J thermocouples, CHROMEL® for type E thermocouples, or with copper for type T thermocouples.

The type "T" thermocouple (copper constantan) is used in applications down to 11 degrees Kelvin (K) (-262 degrees C). Copper of high electrical conductivity and low oxygen content gives a highly reproducible thermoelectric output. Since copper is available in near pure form, it need not be specially selected for thermocouple use as long as it conforms to ASTM specs for soft or annealed bare copper wire. The constantan used with the type J thermocouple cannot be used with the type T thermocouple because of different voltage requirements. A specially formulated constantan must be used. This alloy is commonly known as the "Adams Constantan" alloy. It is a copper-nickel alloy that combined with copper matches the Adams Copper-Constantan Table. The limits of error of the type T thermocouple are totally dependent upon the degree of reproducibility of constantan from melt to melt. Adams constantan is acceptable if it gives emfs against a platinum standard within the limits of error of ±1.5° F. between ­75° and 200° F.

The type "T" thermocouple should not be used above 350° C since the copper will oxidize rapidly above this limit.

The type "K" thermocouple designates any thermocouple which exhibits, within specified limits, the thermal emf characteristics as given in the CHROMEL®-ALUMEL® tables over the range of temperatures from ­253 to 2505° F. The CHROMEL®-ALUMEL® thermocouple is the most common used to meet this criteria.

CHROMEL® is an alloy having the composition Ni90Cr10. Its thermoelectric power against platinum is higher than that of any other commonly used alloy, reaching a maximum of 35 microvolts per degree C. As a nickel-chromium alloy it is resistant to oxidation at high temperature.

The manufacturing processes involved with industrial thermocouple sensors is almost as varied as the applications for which they are intended. Basically speaking, however, industrial thermocouples are made by first forming the measuring (hot) junction by welding the two dissimilar metal wire conductors together. Welding in an inert gas atmosphere prevents oxidation and is highly recommended if the thermocouple is to be used at temperatures exceeding 1000 degrees F. For lower temperature applications, junctions may be silver brazed using a borax flux. The thermocouple wire may be either bare wire conductors or wire conductors housed in metallic sheath material with hard packed powder insulating material usually made of magnesium oxide (MgO). When using the mineral insulated type thermocouple material, the sheath is usually welded over at the same time the junction is welded thereby forming a sealed junction. The bare wire type thermocouple may be housed in a closed end tube or sheath of metal or ceramic after the junction is welded. Before insertion into the well or protection tube, ceramic insulators are placed over the bare thermocouple wires to prevent the shorting of the wire together or to the wall of the tube. The thermocouple assembly is then fitted with the required termination (connectors, connection heads, lead wire, etc.) to complete the assembly.

Resistance Temperature Detectors - History and Theory of Operation

Another widely used device for measuring temperature is the Resistance Temperature Detector or RTD. Fifty years after Seebeck made his discovery concerning thermoelectricity, Sir William Siemens, using research done by Sir Humphrey Davy that determined that the resistivity of metals showed a distinct relationship to temperature change, established the use of platinum as the element of a resistance thermometer or RTD. Platinum RTDs as well as RTDs made from various other metals operate under the principle that the electrical resistance of certain metals increase / decrease in a repeatable manner as temperature increases / decreases.

Resistance temperature detectors are rapidly becoming today's temperature sensor of choice. If fact, there are currently over seventy five manufacturers of industrial RTD sensor assemblies in the United States alone.

Even though RTD sensors tend to be relatively slower in response than thermocouples, Resistance Temperature Detectors offer several advantages over thermocouples as temperature sensors in industrial applications. Typically, at temperatures over 850° C, thermocouples must be used, but for temperatures not exceeding 850° C, RTDs offer a definite alternative. RTDs are usually selected over thermocouples because of their inherent stability. A typical Platinum RTD can be thermally shocked from boiling water to liquid Nitrogen (­195° C) 50 times with a resulting error of less than 2/100's of a degree C. Typical stability is rated at ±0.5° C per year.

Another advantage over thermocouples is that no special compensating leadwire or cold junction compensation is needed.

Briefly, an RTD works like this: Electrical resistance of certain metals increases and decreases in a predictable manner as the temperature increases or decreases. The most commonly used metals for RTDs are Platinum, Copper, and Nickel. There are basically three reasons for selecting these metals over others. First of all, these three metals are available in near pure form. This is important to insure consistency in the manufacturing process. Secondly, these metals offer a very predictable temperature versus resistance relationship. While not perfectly linear, they are much more linear than thermocouples. Also, all three of these metals offer the ability to be processed into extremely fine wire. This is important especially in "wire wound" elements, which are the most common types in use today.

RTD - Types

Among the three metals mentioned above, Platinum is the most commonly used due to its having the best temperature to resistance relationship, its ability to withstand high temperatures, its limited susceptibility to contamination, as well as the best stability. Platinum RTD elements are also the most repeatable and have the broadest measuring range (typically from ­200 to +850 degrees Celsius).

It is noteworthy to mention that Platinum RTDs are used to define the International Practical Temperature Scale (IPTS) from the triple point of Hydrogen (­259.34° C) to the freezing point of Silver (+961.78° C).

Since Platinum RTDs are already used as temperature standards in laboratories throughout the world, advancements in manufacturing techniques that make RTDs more rugged are making RTDs the sensor of choice in many industrial applications.

The typical RTD is constructed by winding a very fine wire of one of the metals mentioned above around an inert substrate such as glass or ceramic material. Most elements are then encapsulated in glass or ceramics to protect them from damage due to vibration, moisture, dust, and migration of foreign metals. The wires from the RTD extend through the encapsulation material where they can then be joined to lead wires by brazing, welding, or soldering. From this point, the RTD element can be housed in tubes or thermowells to match the applications much like thermocouples. (Figure 8)

Most recently introduced to the RTD market has been the "Thin Film" RTD element. This type of element, rather than being "wire wound", is produced through a process known as Thin Film Technology. Developed by the semiconductor industry, this process deposits a thin film of Platinum onto a substrate usually of ceramic material through cathodic atomization or "sputtering". Cathodic atomization works like this: A ceramic substrate made of high purity aluminum oxide is placed in a vacuum opposite a platinum disk. The platinum disk serves as a cathode (carries a negative charge). After the vacuum container is evacuated, a noble gas is introduced and a discharge is fired by means of a high frequency electric field generator. A plasma forms between the ceramic substrate and the platinum cathode. The gas atoms exist in the ionized state in the plasma. The positively charged noble gas ions are accelerated through the electric field toward the platinum cathode. When they strike the surface of the platinum they knock platinum particles off through the force of impact. These particles come off at such a high velocity that they deposit on the surface of the ceramic substrate. Over a calculated period of time, the platinum "sputters" to a layer of defined thickness over the substrate. The layer of Platinum may be as thin as 1 micron. After the deposition is made, a laser is used to trim the platinum layer to a precise resistance. Another even newer technology for producing thin film RTDs is called Thin Film Lithography. In this process, after the substrate has been coated with platinum, a photosensitive lacquer is applied to the platinum layer. The lacquer is then illuminated through a mask with light of a defined wavelength, and then developed. After developing the non illuminated parts of the lacquer are left behind and represent the final design of the RTD conductor. Using a dry etching process, the non covered parts of the platinum coating are removed by bombardment with atoms. The platinum layer beneath the lacquer remains behind on the ceramic substrate. The remaining lacquer is then removed. Finally, the platinum conductors are again laser trimmed to reach the required nominal resistance value. Because of this relatively new technology, RTDs can now be produced in more versatile shapes and designs. They can also be made much smaller than their wire wound counterparts. In fact, it is now possible to manufacture an RTD element the size of a pencil point! You might ask "why make such a small element?" By making the element as small as possible, we can now make an RTD assembly more responsive and "tip" sensitive. The "tip sensitive" nature of a thermocouple has always been an advantage over the RTD. Now this advantage has been all but eliminated. (Figure 8.)

RTDs, unlike thermocouples, are passive devices. They operate as one leg of a bridge network and therefore require a small amount of current, typically one mADC, to produce a measurable resistance change proportional to temperature change. It is important, of course, that the power supply generating the excitation current be stable and that the other legs of the bridge network remain constant with any change in temperature.

There are two popular calibrations of Platinum RTD's in use today. With thermocouples, we work with ISA standards for thermocouple alloys. With RTDs we work with temperature coefficients or "Alphas". The most popular and most used alpha for Platinum RTDs throughout the world is the 100 ohm international or DIN 43760 coefficient of .00385 ohms / ohm / degree Celsius. This means that the element at 0 degrees Celsius has a resistance of 100 ohms while at 100 degrees Celsius the resistance is 138.5 ohms. The other somewhat common but much less popular alpha is the sometimes called American Standard alpha of .003926 ohms / ohm / degree Celsius. The essential difference in the two alphas is that due to a slightly more pure Platinum used in the .003926 alpha, absolute accuracy is slightly better.

A word of warning. Since we are dealing with two alphas, it is most important that our instrumentation be matched to the correct alpha. Just as you would not use a type "J" thermocouple with an instrument calibrated for type "K" thermocouples, you would not use a DIN RTD with an instrument calibrated for a .003926 alpha. Doing so would introduce significant errors into the system.

Care must be taken in ordering both RTDs and the instrumentation associated with them. Just as you cannot assume that because it is a platinum RTD, it has an alpha of .00385, neither can you assume that a specification that calls for an instrument with input for a Platinum RTD requires calibration for the DIN .00385 coefficient.

Although there are advantages to using RTDs over thermocouples, there are some concerns as well. I call them concerns, not problems, because if we understand the applications, the concerns will not become problems.

First of all, there is the concern with "lead resistance". As mentioned earlier, RTDs work as one leg of a bridge. Where RTDs are mounted some distance from the instrument, the most frequently encountered cause for error is with the lead resistance effect. Since the copper conductor in the lead wire can change resistance with temperature, and since these lead wires are in series with the RTD element, an error can be introduced. In order to compensate for these errors, RTDs are often supplied in three or four wire versions. (Figure 9) The extra wire or wires are used to offset the error by balancing the bridge. Because of lead resistance error, three wire RTDs have become the most common RTD in industry with four wire RTDs becoming more and more popular. The only apparent disadvantage of three or four wire RTDs is the need to run extra leads back to the instrumentation. This problem can be avoided if 2 wire transmitters are used.

Another concern is the phenomenon of "self heating". Since current must flow through the sensor, there must be a certain amount of heat energy produced. This additional heat will of course elevate the temperature measurement erroneously. Self heating is expressed as the amount of electrical energy required to raise the output of the sensor by one degree Celsius. It is usually measured in milliwatts. The self heating is usually minimal if the excitation current is kept in the range of one to two milliamps.

An additional concern is the fact that RTD elements are not as rugged as thermocouples. Where necessary (in areas of high vibration or shock), this concern can be lessened by the use of thin film RTDs.

Two, Three, or Four Wires,
That is...

A 2-wire RTD in a typical "Wheatstone Bridge" circuit.

Unlike a 2 wire RTD, a 3-wire RTD will compensate for lead length resistance. An accurate measurement will result only if the length, and resistance of each lead matches exactly (shown; a "Wheatstone Bridge" circuit).

A 4-wire RTD compensates for all resistance imbalances between the leads (shown: a circuit with a constant current source).

Thermistors -

Another resistance based temperature sensor is the Thermistor. Unlike the RTD, the typical thermistor has a negative temperature coefficient. This means that with an increase in temperature, the resistance of the thermistor decreases. Since the per degree resistance change in a thermistor is much greater than with an RTD, a thermistor is quite sensitive to minute changes in temperature. Although the thermistor is a more sensitive device, it is also very non-linear and usually used over a very small temperature span. Thermistors have not gained nearly the popularity of RTDs or even thermocouples in industry due to their limited span as well as other disadvantages. Since thermistors are semiconductor devices, they are quite susceptible to permanent decalibration when exposed to high temperatures. In addition, thermistors are quite fragile and great care must be taken to mount them so that they are not exposed to shock or vibration.

I C Sensors -

Integrated Circuit Temperature Sensors are one of the latest innovations in temperature sensing. The main advantage to this type of sensor is that it is a naturally linear device which provides an output that is proportional to absolute temperature. The output of IC sensors is typically stated in microamps per degree Kelvin. The most common IC sensor in use today is the AD590 manufactured by Analog Devices.

The AD590 acts as a high-impedance, constant current regulator passing 1 micro-amp per degree Kelvin. It uses a supply voltage of between 4 and 30 V.

The IC temperature sensor uses a fundamental property of silicon and germanium transistors, from which it is made, to realize its temperature proportional characteristic. If two identical transistors are operated at a constant ratio of collector circuit densities (r), then the difference in their base-emitter voltages will be (kT/q) linear. Since both k (Boltzmann's constant) and q (the charge of an electron) are constant, the resulting voltage is directly proportional to absolute temperature. This voltage is converted to a current by a low-temperature-coefficient thin-film resistor.

Another popular IC temperature sensor is the LM134/234/334 series from National Semiconductor. This series of ICs make ideal remote temperature sensors due to the fact that they operate on a current output that is unaffected by long wire runs. The output current is directly proportional to absolute temperature in degrees Kelvin. The typical output is 1 microamp per degree Kelvin. Another advantage of this sensor is that it operates on any voltage from 1 to 40 volts DC.

The disadvantages of IC sensors are all of those expressed with RTD's plus a very limited temperature range usually limited to 150° C maximum.

Radiation Sensors -

All of the sensors we have discussed so far have their advantages in certain applications. However, they also share one distinct disadvantage. They must come in contact with the medium who's temperature we desire to monitor. That fact eliminates their usage in very high temperature applications found in many industrial applications. Radiation sensors often offer a solution to this problem.

The radiation sensor can typically measure temperatures up to 3500° C without contact with the measured medium.

Radiation sensors work on the principle that the temperature of a target determines the wavelength of the emitted radiation. The simplest of the radiation sensors is the optical pyrometer, which simply requires the operator to match the color of an incandescent target to a color scale in his line of sight. Other systems are both more complex and accurate. The accuracy of a radiation sensor depends on relative stability in colors, incandescent lighting, ambient temperature of the detector head, detector angle relative to the measured surface, and surface emissivity changes. All of these variables can contribute to system errors. However, progress is being made in the area of radiation sensors. It is now possible by using long wavelength detectors and filters and by using digital sensor head transmitters which output a linear signal, to avoid electrical interference and maintain an accuracy to within a few degrees which is usually adequate in very high temperature applications.

Emissivity is a term for the amount of energy emitting characteristics of different materials. It is the function of wavelength, temperature, and angle of view. Emissivity is defined as the ratio of the energy radiated by an object at a given temperature to the energy emitted by a "blackbody" or perfect IR radiator at the same temperature. Theoretically, a blackbody neither transmits nor reflects energy. The emissivity of a blackbody is represented by "1.0". Blackbodies absorb and re-emit all energy incident upon them and are therefore ideal surfaces for IR measurement. Therefore, blackbodies are used to calibrate IR measuring devices. All objects other than blackbodies have an emissivity of less than 1.0. For example, an object may have an emissivity of .85. That means that the object emits only 85% of the energy emitted by a blackbody. Correction factors and adjustments are usually built in to IR sensors so that they may be calibrated for specific emissivities. If the correction factors are not applied, the temperature reading will be lower than the actual temperature of the object being monitored.

Radiation sensors have their biggest advantage in measuring high temperatures. However, there are many lower temperature applications where non contact temperature measurement is desired.

Small targets and moving objects can be monitored more accurately since IR sensors measure only the energy emitted by the object, not the surrounding area.

Infrared measurements are much faster, allowing several measurements to be made in less than a second. Contact methods require much longer due to slower response times.