Sensors

Sensors are devices that change a parameter into a form that can be measured or recorded easily. In this sense they can be classified as transducers, although not all types of sensors convert one form of energy into another. Thus, there are (true) transducer-type sensors and non-transducer-type sensors. An example of the latter is a mechanical switch operated by a physical movement to “sense” the end of that movement and apply an electrical signal to indicate or control it. It could reverse the direction of current to a dc motor powering the movement, for instance, making the movement re verse when it reached a limiting position. This is consistent with the main use of sensors in providing an electrical signal output that is usually the simplest way of measuring or recording a parameter required.

TEMPERATURE SENSING

Temperature can be sensed and indicated, or recorded, by a thermometer. A simple thermometer, however, can only provide a visual indication of temperature. To convert heat energy (temperature) into a corresponding electrical signal that can be indicated, or recorded, on a meter, you need a time transducer. The most usual type in this case is the thermocouple. Alternatively, heat energy (temperature) can be converted into mechanical energy to operate a pointer movement directly, in which case another type of transducer—the bimetal strip—is used.

Thermocouples are probably the most widely used transducers for temperature sensing. Normally two separate thermocouple functions are used: one for sensing, and one to serve as a reference. If the reference function is kept at a known temperature, then the voltage generated in the circuit is a direct measure of the absolute temperature of the sensing function.

Industrial thermocouples are most commonly made in the form of probes fitting into a protective metal or ceramic tube. There is, however, another very useful type known as a gasket thermocouple. This takes the form of a metal washer to which the thermocouple wires are bonded. This kind of thermocouple can be attached to existing bolts or studs like an ordinary washer, even under spark plugs for measuring cylinder-head temperatures.

THERMISTOR and RESISTANCE PROBES

Another type of transducer used for temperature sensing is the thermistor or resistance probe. Actually there are two distinct types, but they work on the same principle. The material used for the probe undergoes a change in electrical resistance proportional to any change in temperature. Thus, measuring this change of resistance in an electric circuit (as a change in voltage or change in cur rent) gives a direct measure of the change in temperature involved.

In the case of the thermistor and the resistance probe, the change in resistance is opposite in effect. With a thermistor the resistance decreases with increasing temperature, but with a resistance temperature screen the resistance increases with increasing temperature. The reason is that two quite different materials are involved. The thermistor uses a semiconductor material as the sensory element. The resistance temperature sensor uses a metal sensory element.

Thermistor probes are normally in the shape of heads, thin rods, disks, or washers. They are particularly sensitive to temperature changes and are thus capable of providing enough signal strength to operate a meter in a simple circuit without amplification. A typical thermistor probe may show a resistance change from 75,000 ohm at -40° C. to only 40 ohm at 150° C., or something like 400 (delta) per degree °C. temperature change (although not necessarily in strictly linear proportion).

On this basis we could anticipate a typical resistance at 0° C of say, 60,000 ohm, and a resistance of 100° C. of 16,000 ohm. In a simple meter circuit, as in ill. 13-1, with a 15-V supply, this would give a circuit range from 15/60,000 = 0.25 mA to 15/60,000 = 0.93 mA. Thus, a 0-1-mA meter would cover temperature measurement from 0° to 1000 C. comfortably with good scale spacing and no signal amplification necessary.

ill. 13-1. A thermistor in conjunction with a power supply and a milliammeter can be used to measure temperature.

Resistance temperature probes are considerably less sensitive but, being of metal, can be used for much higher temperatures than thermistors. Their upper temperature limit is two or three times higher than the typical limit for thermistors. They are also more suitable for use as sensors of very low temperatures.

The resistance/temperature characteristics of metallic wires are expressed by their temperature coefficient of resistance, which is not a constant. It normally tends to increase with increasing temperature; the change (increase) in resistance becomes more marked. However, some metals shown an opposite effect. Alloys of nickel and manganese, e.g., have equal and opposite nonlinearities in resistance-temperature characteristics. Thus, combining the two in a sensor can provide a substantially linear change of overall resistance with temperature.

Typical performance one might expect from a nickel- manganese sensor is a resistance of the order of 300 ohm at 80° F., falling to about 200 ohm some very low temperature, say around -400° F. The actual resistance change is thus quite small—only about one-fifth of an ohm per degree Fahrenheit.

In practice, resistance temperature sensors are normally based on a single metal of high purity. Platinum is the usual choice because it does not oxidize or corrode (except in the presence of carbon gases) and can be used at high temperatures. Nickel is a cheaper alternative for less exacting commercial duties. Special materials, such as 0.5 atomic-percent iron-rhodium, may be used for measurement of extremely low temperatures. Copper is used for less extreme applications.

It is important that a high-purity wire be used, because the temperature coefficient of resistance of any conductor is sensitive to impurities, most of which depress the value. The purity of platinum required for accurate measurement is extremely high, yielding an alpha value (or temperature coefficient of resistance) for the sensor of not less than 0.003925 between 0° C. and 100° C. The purity of platinum wire is, in fact, normally expressed by its alpha value.

The relationship between variations in resistance of a wire with temperature and the reference scale (gas scale) can be expressed mathematically as

ill. 13-1-1

where tg = temperature on the gas scale tp = temperature on the platinum-wire scale

C = a constant depending on the purity of the wire (for pure platinum the value is 1.5)

Differences between the two scales are simply corrected by calibration.

The typical resistance temperature probe uses platinum, nickel, or copper based in a metal or ceramic enclosure. Because it's a large surface area relative to its mass, it's mainly suitable for measuring temperatures of an area rather than a point.

Such elements are connected to a compensated bridge circuit for signal readout. Figure 13-2 shows a wheatstone bridge circuit (A) and a ratio-arm bridge circuit (B).

CERAMIC TEMPERATURE SENSORS

Certain ceramic materials have an electrical resistance that is largely unaffected by temperature until it reaches a particular value where the material undergoes a crystalline change and the resistance increases abruptly. The temperature at which this occurs is the Curie point, where the resistance rises within a span of a few degrees.

ill. 13-2. Two types of bridge circuits that can be used with thermistors. At A, a Wheatstone bridge; at B, a ratio-arm bridge.

Such ceramic devices thus sense when a particular temperature (its Curie point) is reached, generating a substantial rise in resistance in an electrical circuit. By selecting the composition of the material and its treatment during processing, one can set the Curie point at any level between about 60° C. and 180° C. Also, the resulting resistance change can be controlled, that is, both the actual change in resistance and the small range of temperature (normally not more than 5° C.) over which this change occurs.

For use as a high-temperature warning device, you must specify the Curie point required and know its normal resistance. In the circuit in ill. 13-3 the sensor is connected in series in the power supply to a relay selected and /or adjusted to pull in at a current equal to Vs/Rs, where Vs is the supply voltage and Rs is the sensor resistance. Alternatively, a variable resistor can be incorporated to ad just the current for relay pull-in. When the temperature of the sensor rises to its Curie point, R will increase, causing the relay to drop out. This completes the circuit for the alarm signal, which is held as long as the sensor remains at or above its Curie point.

ill. 13-3. A simple thermistor-operated alarm control for warning of overheating condition.

There are a number of interesting variations in the use of this type of sensor. It can, e.g., be used as an airflow monitor using the same circuit as before. In this case the circuit is set up so that in still air the self-heating effect of the current passing through the sensor causes it to reach its Curie point, presenting a high resistance in the circuit inhibiting pull-in of the relay. In the presence of airflow past the sensor the cooling effect of the air loses its temperature below its Curie point. This results in a substantial reduction in circuit resistance, and the relay is adjusted to pull-in under this condition. Should the airflow cease, the circuit resistance will rise, causing the relay to drop out and activate the alarm circuit.

Another application using the self-heating effect of a ceramic temperature sensor is as a current limiting device in an electrical circuit (ill. 13-4). Here the sensor is merely connected in series as a resistor. Should the current rise above a predetermined level, the sensor will be heated to its Curie point and go into its high-resistance state, automatically reaching the current level in the circuit.

ill. 13-4. A thermistor can be used for limiting the current in a circuit, preventing damage from overheating.

SOLID-STATE TEMPERATURE SENSORS

A simple solid-state temperature sensor can be based on a form of zener diode so constructed that its breakdown voltage is directly proportional to absolute temperature. A typical device of this type may have a working range of -10° to +100° C. with a linear out put of the order of 10 mV per degree Celsius.

Characteristics of such a device will normally specify an out put voltage at a specific temperature; a typical value would be of the order of 3 V at 250 C. This, over the temperature range of, say 0° to 100°C., the anticipated voltage range would be from 2.75 V (at 0° C. to 3.75 V (at 100° C.). Figure 13-5 illustrates a temperature-measurement circuit using such a device.

ill. 13-5. A circuit using a solid-state sensor for measuring temperature.

This device can be expected to have an operating current range from about 0.5 to 5 mA, or an optimum operating current of 1-2 mA. Here the value of the resistor R is selected to give a suitable operating current, established with the supply voltage to be used and the resistance of the device. The latter is normally low enough to be negligible. Thus, for a 10-V supply and a design operating current of 2 mA, say,

R= 10 / 0.002

=5000 Ohm

There are also integrated circuits that include a temperature- saving device of the type described together with a stable voltage reference and an op amp in the same IC chip. If the sensing device has the same characteristics (10 mV per degree Celsius), output voltage from the IC then goes negative at this rate with temperature increase.

The particular attraction of such an IC is that any temperature-scale factor can be realized with external resistors.

PYROMETERS

Pyrometers are another type of temperature sensor. Optical and radiation pyrometers depend on the intensity of the radiation, either monochromatic or total, emitted by a hot body. Optical pyrometers rely on visual observation of the indicator; radiation pyrometers use a receiver, such as a thermocouple, that reacts to the change in temperature produced by the radiation.

The principle of operation is based on the fact that the total energy by a blackbody is proportional to the fourth power of its absolute temperature, or

E = O (T⁴ - T₀⁴)

where E = total energy radiated at absolute temperature T

T = ambient temperature

O is a constant.

Because T is usually small relative to T, this formula simplifies to

E = OT⁴

In the case of an optical pyrometer the intensity of the light emitted by the hot body is compared with that given by a standard hot body in the same wavelength, usually either by matching to a constant comparison lamp or by adjusting the brightness of a single lamp until the filament “disappears” (has the same color temperature as the emitted light). Optical pyrometers are usually fitted with a red monochromatic glass, which assists color matching and also protects the eye. They can be used to measure temperatures from about 100 deg. C. and up but are normally only employed for measuring very high temperatures above the range covered by other instruments (about 2000 deg. C. and above). They are generally less accurate at any temperature than the other types of thermometers described.

In addition to the well-established forms of optical and radiation pyrometers, other instruments include:

Infrared radiation pyrometers: generally more sensitive to lower temperature and can give accurate readings within 1 to 2 percent down to temperatures as low as 50 deg. C.

Two-color pyrometers: operate on two wavelengths used on a ratio basis and eliminate the need for estimating the emissary effect of nonblackbodies (or coating the surface with black).

Color still and cinematography: the optical density or color of the film provides a quantitative measurement of temperature.

Thermographic pyrometers: the ultraviolet content of the radiation is determined by its effect on a phosphor-coated screen, which can be analyzed quantitatively in terms of optical density. Largely, however, this method is mainly useful for displaying the temperature pattern over a surface.

HEAT-SENSITIVE MATERIALS

Finally, on the subject of temperature sensing, a number of materials can act directly as sensors by changing their color or appearance at a specific temperature or specific combination of temperature and exposure time. These are known as thermographic materials.

Simple examples are crayons or lacquer, which can be applied to a surface to produce a mark. Once a particular temperature is reached, the mark liquefies sharply, the change in appearance be ing easily noticeable. Similar materials are also available in pellet form, the “calibrated” temperature being indicated by the first signs of liquid appearing due to melting of the pellet.

Quite close “calibrated” temperatures can be provided; e.g., steps of about 3 to 4 degrees Celsius over a range that may extend from about 45 deg. C. to over 1000 deg. C. are common. Accuracy of temperature rating can be within plus or minus 1 percent.

Temperature-indicating materials that liquefy at a specific temperature are usually thermoplastic or “reversible.” That is, once having melted, they will solidify again to a “dry” mark when the temperature falls below the rated value. Other crayons or lacquers may undergo a permanent color change at their rated temperature and thus don't need monitoring or a continuous watch. They indicate that a specific temperature has been reached or exceeded. A series of marks with different temperature ratings could be used to indicate a peak temperature reached when the subject was not under observation.

Pyrometric cones and bars are another type of heat-sensitive material, mainly used for temperature-time or “heat-soak” indication in pottery kilns (although they can have other industrial applications). They are refractory materials cast in rigid shapes that are maintained until the cone or bar has absorbed a specific quantity of heat. At this point they will sag, and if heating continues, they will melt. A series of such cones, each with a different temperature rating, can thus be used to determine oven or bulk heating by direct observation or to record the amount of heating they have received during a firing cycle.

A variation on this technique is the use of special alloy “plugs” or “pellets,” that undergo a permanent change in hardness, permeability, or some other physical property when exposed for a certain duration at a fixed temperature. After a given time they are removed, and the change in characteristics is measured, from which the degree of heat soaking they have received can be assessed.

LIGHT SENSORS

Light sensors or true optical transducers are based on both photovoltaic and photoconductive devices. Such systems can be extremely flexible, capable of detecting both visible and invisible light signals, and used to detect, count, stop, start, sort, measure, position, control, and perform any number of similar functions.

Light sensors are packaged as individual units or in various arrays for reading multiple-impulse patterns. Complete units with associated optics and electronics may be available as scanners, recorders, counters, or cartridges. In this case four basic elements are involved: a light source, an optional system, a photoelectric sensor, and necessary electrical processing equipment (although, as is common, the latter may be separated from the sensor package).

Light sources may consist of an incandescent lamp, a red lamp, or a LED (emitting visible light or infrared).

The optical system normally employs lenses, mirrors, and prisms, as appropriate, for straight-line optive paths. For more complex paths, or to simplify presentation of light at a point source, fiber-optic bundles can be used.

The photoelectric sensor used is normally one of the four following types, the choice largely depending on the application.

Photovoltaic cells. self-generating with a maximum output of the order of 0.5 V. Output is substantially linear with illumination and varies logarithmically with incident radiation. Types of cells include silicon, selenium, germanium, gallium arsenide, and indium antimonide, with silicon and selenium being the most active types. Silicon cells provide approximately 20 times more output than selenium cells and have a much broader range of frequency response. Selenium cells are more responsive in the visible-light range, although they are now outperformed by second-generation silicon cells. Selenium cells are also subject to fatigue with repeated exposure to high levels of illumination.

Photodiodes: operate in a switching mode.

Phototransistors: also operate in a switching mode but with amplification of signal or gain in proportion to the circuit of light falling on them.

Photo SCRs: capable of switching large amounts of power as the result of the absence or presence of light. This can result in considerably simplified electronic circuitry for on-off (digital) switching. Photo SCRs can switch as much as 500 mA at 200 V, with trigger threshold readily adjusted by bias resistance.

PROXIMITY SENSORS

Proximity sensors detect and /or measure the proximity of some object relative to a base position without actually touching the object. They are thus noncontacting sensors or probes in the mechanical sense. A variety of principles is used in the design of these sensors, including reflected light, radiated heat, reflected electromagnetic radiation (up to and including radar, which is a more extreme form of proximity sensor), pneumatic (airflow) and fluidic devices, and magnetic disturbance. Some are power sensors and some are true transducers.

Proximity sensors are described collectively here. Photoelectric devices are widely used as proximity sensors .

FIBER-OPTIC TACHOMETER

The fiber-optic tachometer consists of a flexible light guide connected to an electro-optic unit. The guide consists of mixed light- transmitting and light-receiving fibers terminating in a focusing optional head. Light from the electro-optical source passes down one set of fibers when it's focused and directed onto a target on the straight edge or rotor to be measured. The target consists of a reflecting strip, so that light is reflected back through the second set of fibers to the detection circuit in the electronic unit.

Each movement of the target past the end of the light probe (once per revolution) thus generates a corresponding return light pulse, which the electronic unit converts into an electronic pulse— usually a direct digital readout signal, the pulse frequency of which is directly related to shaft rotational speed.

This type of transistor has many advantages for revolution counting. Alignment problems are minimal because the transmitter and receiver are in the same head; also, the probe can operate over a wide range of distances from the target (typically from 1/4 in. to over 20 in. if necessary). The output pulse signal is noise free and of constant amplitude, making it easy to process. It can also work over a wide range of frequencies from 0 (100-percent work space pulse on target) to about 6000 pulses per second (equivalent to a rotational speed of 360,000 rpm with a single target). The actual signal-pulse frequency generated can also be adjusted upwards for low speeds (low rpm) by using multiple targets on the shaft or rotor.

PREV: Photoelectric Devices
NEXT: Thermocouples

Fundamentals of Transducers (all articles)