Home | Forum | DAQ Fundamentals | DAQ Hardware | DAQ Software
Input Devices | Data Loggers + Recorders | Books | Links + Resources
We pointed out earlier that most remote control systems and all automation systems must have a device that will sense the status of whatever it's we are controlling. With remote control systems we must also have an indicator that will show us this status. If we are to remotely turn on a light, or open a door, we must first know whether the light is on or off, or whether the door is open or closed.
The function of the sensor is to sense the status of what we are controlling and provide an electrical signal that is a function of the status. The easiest sensors to build are completely electrical ones. These are used when the quantity that we wish to sense is an electrical quantity. In this case, all we have to do is convert from one electrical quantity to another, usually from one voltage level to another.
When the quantity we wish to sense is not electrical, the problem is more complicated. To take a simple example, suppose that we are controlling a door with a device that will automatically open and close it. We need a sensor that will respond to the position of the door and provide an electrical signal that we can use to determine the position of the door. Converting a mechanical quantity to an electrical quantity is usually much more involved than simply changing a voltage level.
It is very important that we know just what information we need from a sensor. If we can get by with a simple on-off indication the problem will be simplified considerably. On the other hand, if we need a continuous indication that is quite accurate, designing and building the sensor can be quite a task.
In any case, we usually want the signal from the sensor to be a low-voltage dc or ac signal that can be handled without any special wiring or insulation.
An electrical sensor is used whenever we want to sense an electrical quantity. We would use such a sensor whenever we want to know whether or not electrical power is being applied to the device we are controlling. Devices of this type include lights, fans, air conditioners, pumps and other electrically operated devices.
A typical example might be where a push button in the living room is used to control an air conditioner or fan in a bedroom in another part of the house. You would probably use the control system to turn on the air conditioner say half an hour or so before bedtime on hot summer nights. The sensor and indicator are needed so you will know whether or not the air conditioner is turned on. With several people living in a house anyone may think that some one else has turned on the air conditioner. The indicator will show for certain whether the air conditioner is on or off.
Figure 2-1 shows a simple arrangement that can be used to sense whether or not power is being applied to an appliance. In ill. 2-1A, the primary of a step-down transformer is connected across the power line feeding the appliance we are controlling. The low voltage from the secondary is fed back to the control point to show us when the appliance is energized. All that this voltage will be used for is to operate a small indicator lamp, so a very small transformer can be used. The only reason for using the transformer in the first place is to isolate the indicator from the power line so that we can use a low signaling voltage.
In ill. 2-1B, we have added a rectifier to our sensing transformer. This simply converts our sensing signal from ac to dc. About the only time we would need to do this would be if our signal was to operate a dc relay. We might also wish to use dc indicating signals if the signal wires were run in the same cable with intercom wires that might otherwise pick up hum.
Figure 2-2 shows another power sensing system. Here an ac relay is connected across the power line feeding the appliance that is remotely controlled. A relay has some advantages over the transformer of ill. 2-1. With a relay as a sensor, we can use any type of sensing signal that we wish, and it's easy to get both an on and an off indication.
Referring to ill. 2-2, we can run three wires from the sensing relay to the control point. When the circuit is closed between lines A and B, it means that the appliance is on. When the circuit is closed between lines B and C, it means that the appliance is off.
In applications where the device being controlled is a long distance from the control point, running three wires as in ill. 2-2 can be expensive. By adding four diodes to the circuit we can do the same thing with only two wires. Figure 2-3 shows the arrangement; here our sensing relay is used to connect one of two diodes into the circuit. When the armature of the relay is down, diode D1 is connected between the two sensing lines. When the armature is up, diode D2 is connected between the two sensing lines. This, in effect, connects a diode between the lines in one direction when the appliance is on, and in the other direction when the appliance is off.
This sensor can be used with the indicator circuit shown in ill. 2-4. Lamps L1 and L2 can be labeled ON and OFF respectively or different colored jewels may be used to indicate ON and OFF. The on and off conditions can be recognized for a considerably distance. A single transformer with a twelve-volt secondary is used as a power supply. The coil of relay K1 is connected across the appliance we are controlling. Consider first the situation where the controlled device is on and diode D1 is connected in the circuit. During the half cycle when the ungrounded side of the transformer secondary, point A in the figure, is negative with respect to ground no current will flow through either of the lamps because diode D1 will be reverse biased. Now during the opposite half cycle when point A is positive with respect to ground there will be a complete circuit through diodes D1 and D3 and through lamp L1.
When the controlled device is off and diode D2 is connected in the circuit by the relay the situation is somewhat similar. When point A is positive with respect to ground there will be no complete circuit. When point A is negative with respect to ground there will be a complete circuit through lamp L2 and diodes D4 and D2.
The diodes used in this circuit are the garden-variety silicon power supply diodes. About the only specification of interest is the reverse bias voltage rating. This should be 25 V or greater. These diodes are very inexpensive and the features of the circuit make it worthwhile.
In the figure, we have specified 12-volt supplies and lamps. There is no particular reason why we have to adhere to this voltage. Any voltage between about 6 V and 24 V can be used. The lamps must be rated for use at the power supply voltage.
Another good reason for using relays as sensing devices is that it's not necessary to have a voltage on the sensing lines all of the time. The signal, if we can call it that, from the relay is basically a contact closure. A power supply is only used at the indicator at the control point. No power is needed at the sensor. it's possible to use a single indicating arrangement to tell us the status of several different devices. Figure 2-5 shows a circuit for this purpose. Here there are several different sets of input lines. Each of these comes from a sensor circuit like that shown in ill. 2-3. These sensor circuits are connected, one at a time, to an indicator circuit like that used in ill. 2-4. As the selector switch is turned to the different positions, either the ON lamp or the OFF lamp will light showing the status of the particular sensor.
This arrangement makes it possible to use a very small control box to control a large number of different devices. The indicating arrangement can also be used with one of the control circuits to be described later. The control circuitry can be connected to another gang on the selector switch. Thus the control box will have only two lamps, a selector switch and a couple of push-button switches associated with the control circuitry.
The sensing arrangements described above will handle just about any sensing problem where the quantity to be sensed is a voltage.
MECHANICAL POSITION SENSORS and INDICATORS
There are many control situations where the thing that we wish to sense is not an electrical quantity at all, but the position of something. e.g., we may need a remote indication of the position of a door, window, or drapery.
In many applications, the mechanical sensing problem can be simplified considerably because we don't need a continuous indication of position, but only whether or not an object is in a particular position. e.g., all that we might need to know is whether a door is fully closed or not. If it's open, we may not need to know just how far it's open. Or maybe we only have to know the position of something in a sort of rough manner. If this is the case, we can use switches as sensors. This is much simpler than using a sensing device that will give a continuous indication of the position of something. There are many commercially available switches that can be used for this purpose, or you can build your own.
Some of the switches that are commonly used for position sensing are shown in ill. 2-6. Figure 2-6A shows a magnetic door switch. The switch comes in two parts. One part, containing the switch mechanism is mounted on the door frame. The other part containing a small permanent magnet is mounted on the door itself.
The actual switch mechanism is a magnetically operated reed. When there is a magnetic field in the vicinity of the reed, the switch will close. When the magnetic field is removed, the switch will open. The switch is mounted in such a way that when the door is closed, the magnet will be near the switch causing it to close. When the door is opened, the magnet will move away causing the switch to open. These switches are available in either normally open or normally closed configurations. You can choose the configuration best suited to a particular application.
Figure 2-6B shows what is usually called a snap-action switch. The switch is equipped with a small lever that touches the object that we wish to sense. When the lever is moved, the switch is actuated. Like the magnetic switch, the snap-action switch is avail able in either a normally-open or normally-closed configuration. Snap-action switches are also available in more complex configurations such as SPDT and DPST.
Snap-action switches can be obtained with many different lever arrangements so that they can be adapted to different applications. As a rule they require only a very small force or a very small displacement or both to actuate them. Several different snap-action switches can be placed so that they will indicate several different positions of an object.
Figure 2-6C shows a mercury switch. This switch consists of two electrodes, which don't touch each other, and a small pool of mercury. When the switch is positioned so that the mercury contacts both electrodes, the circuit is closed. Mercury switches are very handy for indicating when something is tilted.
CONTINUOUS POSITION SENSORS
The switches described above will usually indicate only one position of an object. e.g., a magnetic door switch will indicate if a door is open, but it won’t tell us how far open. A switch such as this is adequate for many control situations.
There are some control applications where we need to know the exact position of some object. Again, a door is a good example. The most obvious sensor to use for such an application is a potentiometer. To use a potentiometer for an application like this we need a mechanical arrangement that will translate the motion of the door into a rotary motion.
Figure 2-7 shows a potentiometer with a pulley on its shaft. The pulley is turned by a cord which is attached to the door. A weight on the free end of the cord keeps the cord tight. When the system is properly set up, the arm of the potentiometer will be at one extreme of its range when the door is fully opened and at the other extreme when the door is fully closed.
With a source of voltage connected across the potentiometer, the voltage appearing at the arm will be a function of the position of the door. If a linear potentiometer is used, we can connect a meter as shown to indicate the position of the door. If you wish, the scale of the meter can be marked to show the door position.
The potentiometer sensor can, of course, be used with objects other than doors. It could, e.g., he used to indicate the position of a damper in a ventilating duct.
Theoretically, the potentiometer is an excellent sensor. In practice, it has many limitations. The mechanical linkage is usually the problem. Unless great care is taken in design and installation the cord may slip or break and the arrangement may require a great deal of attention. If we don’t need to know the exact position of an object, but can get by with an approximate indication we can use a much simpler arrangement.
The mechanical portion of the sensor shown in ill. 2-8 is somewhat similar to that of ill. 2-7. Here again we use a cord to get a linear mechanical motion that is a function of the position of a door. Here the similarity ceases. We don’t need a potentiometer or a pulley.
The arrangement of ill. 2-8 operates on the principle that the voltage drop across a forward-biased diode is reasonably constant. With a silicon diode of the type used as a rectifier in a power supply, this voltage is about 0.7 V. In our circuit, the cord which is fastened to the door is tied to a flexible wire. Phosphor bronze wire of the type used to hang pictures is fine for the purpose.
The cord and the wire pass through several eyelets. Between each pair of eyelets we have connected two silicon diodes. The diode string is connected through a dropping resistor to a voltage source. A line from the top of the diode string is connected to a voltmeter at the control point of the system as shown in ill. 2-8.
When the door is closed, the wire that is tied to the cord shorts out all of the diodes. Thus the voltage across the string is zero and the voltmeter will indicate zero. When the door is partially opened, the cord, which is an insulator, passes between two of the eyelets. The two diodes connected between these eyelets are no longer shorted and the voltage across the diode string will be equal to two diode-drops—about 1.4 V.
As the door is opened more, more of the diodes will be switched into the circuit. When the door is fully opened the voltage across the string of diodes will be equal to eight diode-drops—about 5.6 V. If you adjust the variable resistor so that you get full-scale indication of the voltmeter when the door is fully opened, the deflection of the voltmeter will be proportional to the amount that the door is opened.
In our example, we used four sets of diodes. We used two diodes in each position rather than one simply to increase the signal to the voltmeter. We could calibrate the scale of the voltmeter to read “Closed—1/4—½—¾—Open.”
There is no reason why the system should be restricted to four steps. If we need more steps, we can simply add more diodes and more eyelets.
When the object whose position we are sensing tilts, we can use the same arrangement with mercury switches across the diodes rather than eyelets. A typical application of this type involves sensing the position of an overhead garage door. When a door like this opens, first the top panel, then the others, tilt from a vertical to a horizontal position.
You can use mercury switches with this kind of a door by means of the arrangement shown in ill. 2-9. A mercury switch is connected across one or two diodes and the switch is connected to an ordinary terminal strip. The terminal strip is attached to the panel of the door and the angle at which the switch is mounted is set so that the switch will open when the door panel starts to tilt.
You can use as many of the arrangements of ill. 2-9 that you wish to get as much resolution of the door position that you need for a particular system.
Naturally, these switching arrangements are not restricted to sensing door positions. They can be used wherever the motion of the object you are sensing can be translated into a tilting motion.
THE VOLTAGE COMPARATOR
A voltage comparator is an integrated circuit that is used to compare the magnitudes of two voltages. It can be used with light-emitting diodes (LED) to provide a graphic indication of the position of anything that can open and close switches. As shown in ill. 2-10, a voltage comparator consists essentially of a very high gain amplifier with two inputs. The output pin is connected to the collector of an npn transistor, the emitter of which is grounded. The gain of the amplifier is so high that the output transistor is either saturated or open. Thus the comparator can be thought of a switch that is controlled by the input voltages.
When the voltage applied to the pin marked + is higher than the voltage applied to the pin marked —, the output will be open. When the voltage applied to the pin marked — is higher, the output pin will be grounded.
In operation, a known reference voltage can be applied to the + input pin and an unknown voltage can be applied to the — pin. When the unknown voltage is lower than the reference voltage, the output pin will be open. When the unknown voltage is higher than the reference voltage, the output pin will be grounded.
The type 339 integrated circuit contains four separate voltage comparators and is called a quad comparator. Figure 2-11 shows the pin connections and maximum ratings of this circuit. The price is quite low and the device is well suited to drive LEDs in a position indicator.
GRAPHIC POSITION INDICATOR
Figure 2-12 shows the diagram of a graphic position indicator using the type 339 quad comparator. The sensor is the same as that used in ill. 2-9 with one additional pair of diodes. A voltage that is a measure of the position of something such as a garage door is generated by switching diodes in and out of a circuit. The circuit of ill. 2-12 uses the four voltage comparators in a quad comparator such as the type 339. The reference voltages are generated from a string of diodes that is nearly identical to the diode string used in the sensor. These reference voltages are applied to the + inputs of the comparators. The — inputs are connected in parallel to the voltage generated by the sensor.
To understand how the circuit works let’s start with the situation where all of the switches in the sensor are closed. Under this condition, all of the comparator — inputs are less than the reference voltages. Thus the outputs of the comparators will be high. That is, the output pins will be open circuits and no current will flow through the LEDs. Now let’s consider the situation when the bottom switch in the sensor portion of the circuit is open. Now the voltage from the sensor will exceed the reference voltage applied to the bottom comparator in the figure. The pre sense of the extra diode at the bottom of the sensor circuit will assure this. The output of the bottom comparator will now be grounded and current will flow through the LED that is connected to it. Now when the next switch in the sensor opens a similar situation will prevail at the second comparator from the bottom and its LED will light. Thus the string of LEDs will light up starting at the bottom of the figure as the corresponding switches in the sensor are opened.
The LEDs can be arranged on a panel to give a graphic display of the position of the object that is connected to the sensor. e.g., if the sensor switches are mercury switches attached to the panels of an overhead garage door, the diodes can be arranged in a vertical line as shown in ill. 2-13A. When the door is fully closed, all of the LEDs will be out. When the door is about one quarter open, the bottom LED will light. When the door is half open, the bottom two LEDs will light. When the door is three quarters open, the bottom three LEDs will light and when the door is fully open all of the LEDs will light. Thus the display will graphically resemble the position of the door. If the sensor is used with an ordinary door, the LEDs can be arranged in an arc as shown in ill. 2-13B to provide a graphic representation of the position of the door.
In ill. 2-12, we used four comparators and four LEDs. This is probably enough for almost any position indication required in a home control system. It can be expanded, however, by adding four more comparators, giving eight increments of indication. In ill. 2-8, we used two silicon diodes for each switch position. This gives us voltage increments of about 1.4 volts. This really isn’t necessary. We could use one diode across each sensing switch giving a voltage of about 0.7 volt. The only reason for using a greater voltage increment is that the circuit is easier to adjust and is less susceptible to noise.
The circuit of ill. 2-12 isn’t very critical. The two resistors used at the + input of each comparator provide a little positive feedback that will make the comparator switch very rapidly from the on to the off condition and vice versa. This will keep the comparators from oscillating even though the gain of the amplifiers is very high. The diodes are not at all critical. About the only requirement is that they all be silicon diodes to provide the constant voltage drop of 0.7 V and that the inverse voltage rating be about 25 volts or greater. The LEDs are not critical either. Just about any LED can be used in the circuit. If the current rating of the LED is very small, the resistors in series with them can be increased in value. Higher current LEDs can also be used by decreasing the value of the series resistors. Each comparator will ground a current of about 15 mA.
One of the things that you might want to sense in a control system is light. Many control systems will operate in one mode during daylight hours and in another mode after dark. e.g., there is no need for a system to automatically turn on outside lights if it's daylight anyway. On the other hand, it might be desirable for the system to turn lights on after dark.
Another place where a light sensor may be useful is to sense the position of some object in a situation where it's practically impossible to make any physical contact with the object itself. This would be the case with a continuously revolving object.
One of the handiest light sensing devices available to the experimenter is the photovoltaic solar cell. Because of the recent energy shortage much attention has been paid to the various ways to convert sunlight into electricity. This development, together with efforts in connection with the space program, has led to the production of many low priced solar cells.
Figure 2-14 shows a sketch of a typical selenium solar cell. it's a little more than one-inch square and has two flexible wire leads. One side of the cell is photosensitive. The solar cell is rugged in construction and provides a large enough output so that it's not particularly sensitive to noise.
The solar cell produces an output voltage of about 0.5 V. This voltage is nearly constant regardless of the amount of light reaching the cell. Once a little light reaches the cell, the output will rise to about its final value. The output current on the other hand, depends on the amount of light reaching the cell.
Inasmuch as the solar cell is basically a current generator, it's ideal for use with a transistor which is basically a current amplifier. Figure 2-15 shows the circuit of a light sensitive relay using a solar cell. Note particularly that a germanium transistor must be used in this circuit. The base to emitter voltage of a germanium transistor is only about 0.2 V, so the voltage from the solar cell is high enough to make base current flow. A silicon transistor has a base to emitter voltage of about 0.7 V and the only way that it will work with a solar cell is if two or more cells are connected in series to get enough voltage to allow base current to flow.
The sensitivity of the circuit of ill. 2-15 is set by means of the 5 K pot, R1. Just about any small signal germanium transistor will work in the circuit. The sensitivity should be set so that the relay will operate at the desired light level.
The circuit in ill. 2-16 can be used as a very handy light meter for setting up control systems that respond to changes in light level. This circuit uses a 0 to 1.0 milliammeter as an indicating device. For control purposes, the light meter can be compared with the photo relay of ill. 2-15. The indication at which the relay operates can be noted. Once this has been done, the light meter can be used to check any location to see if there is enough light to operate the photoelectric relay.
The photo relay of ill. 2-15 can also be used to sense the position of an object. Photoelectric sensing is particularly useful with objects where it's very difficult to arrange a switch that will be opened or closed by the object. There are some garage doors that are suspended in such a way that it's nearly impossible to use a switch to sense the position of the door without interfering with its motion. Another application might be where it was desired to turn on a light whenever anyone entered an area through a rather large passageway.
The arrangement of ill. 2-17 may be used for sensing the position of an object. The light source is an ordinary panel lamp that may be powered either by a battery or by a small step-down transformer connected to the power line. The lamp is mounted inside a small cardboard tube merely to prevent the light from being distracting. If the distance between the light source and the photo relay isn’t very great, it will usually not be necessary to focus the light. If focusing is necessary, the simplest way to accomplish it's to use the reflector from a flashlight. The photo relay uses the circuit of ill. 2-15. The solar cell and the electronic circuit are housed in a small cardboard tube. The reason for the tube is to keep ambient light from reaching the solar cell and providing a false signal. If the ambient light level is high, it will help to paint the inside of the cardboard tube black.
There are instances where the object whose position is to be sensed is not only inaccessible as far as mounting a switch is concerned, but it may be nearly impossible to arrange a photo relay so that the object will interrupt a beam of light when it moves. Figure 2-18 shows a damper that is mounted inside a ventilator. In this case, it would be nearly impossible to arrange a switch so that the damper would contact it, and it would be almost as difficult to arrange a photo relay so that the damper would interrupt a light beam. In the figure, the problem is solved by cementing a small mirror to the side of the damper. When the damper is closed, light from the light source will be reflected off the mirror and will hit the photo relay. As soon as the damper starts to open, the mirror will move so that the light will no longer hit the photo relay.
Arrangements of this type are limited only by the ingenuity of the experimenter. With the proper arrangement of mirrors, it's possible to sense the position of almost any object, no matter how inaccessible it might be.
All of the sensors described so far depend on a contact closure or opening to sense the position of the object that is to be controlled. In most home control systems, this will be entirely satisfactory. Contacts are reliable, and when anything fails in the system it's very easy to troubleshoot. One major limitation of the use of switches and contacts as sensors is that a separate wire is required for each sensor. In many simple applications, this isn’t much of a limitation. Small wires can be used for the sensing and they are not very expensive. However, this limitation demands consideration if a large number of sensors are used, and /or if they are located a substantial distance from the main control point of the system. In such a case it might be worthwhile to multiplex the lines from the sensors. One way of multiplexing these signals will be described shortly.
In a number of applications, the simple on/off sensing of the circuits described so far will not be adequate. A number of photo sensitive devices are available to provide continuous control over a range of lighting levels.
Figure 2-19 shows the schematic symbol for a photoresistor. As the name suggests, this component is essentially a light con trolled potentiometer. The amount of light striking the sensor surface will determine the resistance of the device. A photoresistor can be used in place of a standard mechanical potentiometer in virtually any circuit.
Photodiodes (ill. 2-20A) and phototransistors (ill. 2-20B) are also available to the experimenter. In a phototransistor, the amount of light striking the sensor serves the same function as the signal applied to a base of an ordinary bipolar transistor.
Light signals can also be used to interface separate circuits. This can ensure maximum isolation between circuits. An optoisolator consists of a light source (usually an LED) and a light detector (photoresistor, photodiode, or phototransistor) within a light tight package.
A signal fed to the input controls the amount of light generated by the LED. This light is sensed by the photodetector and produces the appropriate signal at the output with complete electrical isolation between the input and the output. You can make your own optoisolator with separate LEDs and light sensors and a home-brew light tight housing.
Lengthy connecting lines can be transmitted by light through optical fiber cables with an LED at one end and a photodetector at the other. The input and output devices must be firmly attached to the fiberoptic cable with a light tight connection to minimize leakage. A fiber optic cable can be bent and twisted as necessary without significant loss of the transmitted light signal. In most home applications, standard wiring will be sufficient, but you might want to consider fiber optics for special purposes.
A photodetector can also be employed as a motion detector. The D1072 is an LSI (large scale integration) device designed for dedicated use as a motion detector. The block diagram for this device is shown in ill. 2-21.
The built-in sensor for this device is a simple photodiode (D1). The rest of the circuitry is set up to convert the output of the photodiode for motion detection. The exact amount of light falling on the photodiode sensor is basically irrelevant. Changes in the light level (such as may occur if an object moves within the sensor’s range) are detected by this circuit.
Many other motion detectors that have been designed require some kind of light transmitter and receiver that generally must be carefully aligned. The D1072 functions in whatever ambient light is available. Transmitter/receiver motion detectors usually require that the transmitted light beam be completely broken for sure detection. The D1072 motion detector can respond to changes in the light level as small as ±5 percent.
Since this sensor will respond to increases as well as decreases in the light level, it can't be confused by reflective surfaces.
The D1072 motion detector can be used under light levels varying over a 1000:1 range. The ambient light may be anywhere from 0.1 to 100 candlepower. This range should be more than adequate for virtually all applications.
The LSI circuitry of this chip is packaged in a clear plastic DIP housing, with a molded lens mounted over the photodiode. This lens improves the detector’s sensitivity at low light levels. Light is gathered by the lens in such a way that a 2-foot circle up to 8 feet away from the sensor can be monitored.
A typical motion detector alarm circuit built around the D1072 is illustrated in ill. 2-22. Any light reaching the internal photodiode (D1) generates a small voltage. This voltage will vary in step with any fluctuations in the amount of light striking the sensor.
Capacitor C5 (connected to pins 6 and 7) couples this changing voltage to amplifiers A2 and A3. Capacitors C5 and C3 (pin 9) serve as part of a filter network to allow the circuit to respond to low frequency voltage changes (light fluctuations due to motion of an object within the monitored area), but constant voltages (unchanging light levels) are ignored.
High frequencies are attenuated by C4 (pin 8) so the unit will not be falsely triggered by normal flicker from ac powered lamps.
Al through A3 precondition the signal to produce a trigger signal for the detector stage. When the voltage (light level) changes more than five percent in either direction, the detector will be triggered, producing a brief burst of tone from the IC’s internal digital generator. The nominal pitch and the rate of pitch change are determined by the values of C1 (pin 12) and C2 (pin 10). The speaker will generate a whooping alarm sound for about 4 to 12 seconds whenever any motion is detected by the sensor.
Instead of the audible alarm, the output of the D1072 motion detector could be used to drive almost any electrically controllable device.
The Taguchi Gas Sensor, or TGS, can be used to detect the presence and degree of concentration of a wide variety of deoxidizing gases, such as those listed in Table 2-1.
This device is extremely sensitive to gas concentrations over a wide range. Depending on the external circuitry used with the TGS, it can be used to measure minute concentrations of just a few ppm (parts per million), or the high concentrations that might be found in specialized environments, such as within an automobile’s exhaust pipe.
The Taguchi Gas Sensor is a resistance type sensor, acting as a variable resistor whose value is set by the level of concentration of the detected gas.
Automated safety systems are a natural application for the TGS. When a higher than acceptable concentration of a toxic gas is detected, an exhaust fan could be turned on to increase ventilation within the protected area.
Table 2-1. Deoxidizing Gases Detected by the Taguchi Gas Detector:
Another application would be automatic adjustment of a burning fuel mixture to maintain concentrations within a specific range. This can be a good way to maximize fuel burning efficiency.
HALL EFFECT MAGNETIC SENSOR
If a current flows through a conductor or semiconductor under the influence of a magnetic field at right angles to the direction of current flow, a voltage drop will be produced. This is the Hall effect.
Hall effect magnetic sensors can prove useful in a number of automation and remote control applications where a magnetic field exists. Any inductor (or coil or transformer) produces a magnetic field. This is also true of motors.
Sprague Electric Company, e.g., manufactures several Hall effect magnetic sensors in IC form. A block diagram for the UGN-3020T Hall effect switch is shown in ill. 2-23. The pinout diagram for this device is given in ill. 2-24. The amplifier boosts the voltage from the actual Hall effect sensor. When the amplifier’s output exceeds a specified threshold level, the Schmitt trigger stage turns on the output transistor.
Output oscillations are prevented by the hysteresis of the Schmitt trigger stage. If the strength of the monitored magnetic field is near the critical threshold level, it may wobble back and forth on either side of the triggering point of the Schmitt trigger. This could cause the output to be erratically switched on and off. The hysteresis built into the Schmitt trigger requires the amplifier’s output voltage to drop significantly below the turn on level before the Schmitt trigger will snap back off. This greatly improves the stability of the device.
Physical (mechanical) pressure along the y-axis of a crystal will generate a voltage across the x-axis. This is known as the piezoelectric effect. The usefulness of this effect in automation and remote control systems should be fairly obvious.
A variation on this basic principle is employed in hybrid piezoresistive IC pressure transducers. These devices are found in many modem pressure sensing applications in place of older mechanical type pressure sensors.
These hybrid IC sensors are smaller and more reliable than their mechanical crystal counterparts. Moreover, they are virtually insensitive to mechanical vibration and offer frequency responses that allow operation right up through the audio frequencies.
Piezoresistive IC pressure transducers can be designed for a wide variety of pressure ranges. Units are available for measuring pressures from 0 to 5000 psi (pounds per square inch).
A typical device of this type is the LX17O1 absolute pressure sensor from National Semiconductor. This IC is referenced to 0 psi (i.e., a vacuum—no pressure). It can measure pressures over a 10 to 20 psia (pounds per square inch, absolute) range. Normal atmospheric pressure is typically about 15 psia, so it's comfortably within the range of this sensor.
This unit is remarkably stable, with a stability rating of ±0.05 psi, and a temperature coefficient of ±0.0054 psi per °C.
Another ingenious pressure sensor that has appeared in recent years is a special pressure sensitive paint made by ELAB Microducers. This paint can be applied over almost any surface to create extremely simple, inexpensive, and reliable pressure sensors for virtually any special purpose application.
Mechanical pressure sensors are not always built around quartz crystals and the piezoelectric effect. A spiral coil of hollow glass, metal, or quartz can be used as a pressure sensor if one end is sealed. The coil will slightly wind tighter or unwind as the pressure of a liquid or gas within the hollow tube is varied.
Air pressure switches are also available. Such a switch can be closed by air pressures as low as 0.02 psi. This is about what might be felt from a gentle puff of air from a few inches away.
Typical applications for air pressure switches include the following:
This type of device can directly drive a LED indicator or other low current circuit. To control circuitry with higher current requirements, the output of the air pressure switch can be applied to a relay, of SCR.
Temperature sensors can also be extremely useful in automation and remote control systems. They can be used to monitor and control furnaces, air conditioners, stoves, and other such equipment.
The simplest type of temperature sensor is the thermocouple. This is nothing more than a junction of two dissimilar metals. If such a junction is heated, a voltage which is proportional to the temperature of the junction will be developed across the two wires. This is called the Seebeck effect, and is created by the different work’ functions of the two metals.
A similar approach is used in standard mechanical thermostats. Two dissimilar metals are placed back to back. The difference in the rates of expansion and contraction due to changes in temperature causes the two-metal sandwich to bend back and forth, opening and closing switch contacts.
Semiconductors are quite sensitive to heat, and can be used for temperature sensing. By applying a small forward bias to an ordinary silicon diode, the temperature can be determined by measuring the voltage drop across the diode. This voltage drop will typically change about 1.25 millivolts (0.00125 volt) per °F. For Celsius measurement, the voltage drop is approximately 2.24 mV (0.00224 volt) per degree °C.
Bipolar transistors make even better temperature sensors than diodes, especially if they are diode connected, as illustrated in ill. 2-25. The base/emitter voltage is dependent on the collector current and the temperature. By holding the collector current constant, we can measure the temperature by monitoring the voltage across the base and the emitter of the transistor.
A specialized, junctionless semiconductor device for sensing temperature is the thermistor, or THERMal resISTOR. This component’s resistance varies with changes in temperature.
There are two basic types of thermistors. A positive coefficient thermistor’s resistance increases as the temperature rises. Conversely, in a negative coefficient thermistor, increases in the temperature will result in a lower resistance.
One problem with thermistors is that their response is not linear, making direct measurement techniques inefficient for temperature measurement. They are great for use over a relatively limited range. For a wide range thermometer, you would need to add some external circuitry to linearize the response of the device.
Later in this guide we will describe a method of using audio tones of various frequencies to initiate control action. This same system of tone generators may also be used to multiplex the signals from many sensors onto a single line. We will not describe the details of the tone generators and decoders here, but will merely show how the principle may be applied to the sensor multiplexing problem.
Figure 2-26 shows a block diagram of a tone multiplexing system that puts the signals from three separate sensors on a single pair of wires. There is a separate tone generator at the location of each sensor. Suppose, e.g., that the frequency of the generator at sensors 1, 2, and 3 is 1000 Hz, 2000 Hz, and 3000 Hz respectively. The outputs of all three of the tone generators are connected in parallel across the signal wires.
At the main control point of the system there are three decoders, also connected in parallel across the signal wires. The first decoder will have an output whenever there is a 1000 Hz signal on the line. The second decoder will respond to a 2000 Hz tone and the third will respond to a 3000 Hz tone. When any of the sensor switches is closed, the corresponding tone will be placed on the signal line. Thus, if sensor switch 1 is closed there will be a 1000 Hz tone on the line. This tone will be sensed by decoder 1 which will produce an output. The other two sensor switches work in the same way.
The tone multiplexing system has the disadvantage that a separate tone generator and its power supply are required at each sensor location. If the distance between the sensors and the control point is great enough, this may be worthwhile. In most cases, it's easier and will cause fewer problems to simply run a separate wire from each sensor back to the control point of the system.
Figure 2-27 shows a schematic diagram of a system that can be used to give a remote indication of the position of anything. The proper name for the system is a synchro system, but in the past it has also been called a selsyn (an acronym for self synchronous) system.
The system consists of two devices, called synchros, which look very much like small electric motors. The generator and motor are nearly identical in construction. In each case, the stator has three windings that are spaced 120 degrees apart. The rotor has a single winding.
Referring to ill. 2-27, it's obvious that when an ac voltage is applied to the rotor of the generator, voltages will be induced in the stator windings. The magnitudes and phases of these induced volt ages depend on the angular position of the rotor. T of the generator and the motor are connected in parallel across an ac source. Note that both rotors are in the same angular position. Under this condition, the voltages induced in the stator of the motor will be the same as those induced in the stator of the generator. As a result, no current will flow in the three wires connecting the two stators. Furthermore, there will be no torque on either rotor.
Now suppose that we turn the rotor of the generator. The voltages induced in the stators of the generator and the motor will no longer be equal. Currents will now flow in the three wires connecting the motor and generator stators. These stator currents will produce magnetic fields. In the motor, the field will interact with the field from the rotor in such a way as to make the rotor turn. If there isn’t much restraining force on the motor shaft, it will turn in the same direction that we turned the generator shaft. The motor will continue to turn until its rotor takes the same angular position as the rotor of the generator. Thus a synchro system consists of two small devices that resemble electric motors and are connected together by five wires. When the shaft of the generator is turned, the shaft of the motor will follow it.
Figure 2-28 shows an application of a synchro system. The generator is connected through a cord to an overhead garage door. When the door opens or closes, the cord will cause the shaft of the synchro generator to turn. Say, e.g., that when the door moves from fully open to fully closed, the shaft of the generator makes 10 complete revolutions.
The synchro generator is connected through five wires (as in ill. 2-27) to a synchro motor at the control point. The shaft of the motor is connected through a 40 to 1 gearing system to a pointer. When the door moves from fully closed to fully open, the generator will make 10 complete revolutions and the motor will follow it. Because of the 40 to 1 gear ratio, the pointer attached to the synchro motor shaft will move through an angle of 45 degrees. A scale can be provided to show just how far open the door is at any time.
The synchro system has the advantage that it can be used to provide a precise, continuous indication of the mechanical position of anything that can be attached to the generator shaft. Because of its construction, a synchro has little wear and will certainly have a much longer life than a potentiometer used in a similar application.
The synchro system also has many disadvantages. In the first place, we need five wires to transmit the position information. Probably more important, we need the full rotor voltage, often 120 V, at the control point. This defeats the advantage of using low voltages as we have in most of the other arrangements described in this guide. Fortunately, few control systems require a sensor with the accuracy and precision of a synchro system.
New synchro motors and generators are probably too expensive for use in a home control system. Many different synchro devices are available in salvage stores that handle military and industrial surplus items. Many of these synchros are rated at 120 V, 60 Hz and can be used directly.
Some synchros, particularly those used in aircraft applications, are rated for use on 400 Hz lines. A unit of this type will burn out if it's used on a 60Hz line; however, some units will function at 60Hz at reduced voltage.
A MOISTURE SENSOR
One quantity that can't conveniently be sensed with any of the arrangements that we have considered so far is moisture. It will not operate a switch and doesn’t lend itself to operation of either photo relays or synchros. Moisture sensing is a valuable control system function in many applications. An automatic lawn sprinkler might be actuated by a sensor that responds to the amount of moisture in the earth. A storage area might have a moisture sensor that would detect dampness and automatically turn on a heater that would dry the area.
A good way to sense moisture is to have two electrodes that are spaced very close together. Figure 2-29 shows such a sensor. it's made from a copper plated printed circuit. The foil is either etched or scraped away so that there is a very narrow space between the two conductors. The amount of space that should be left between the two plated electrodes depends of the desired sensitivity of the detector and the type of material that the board is made of.
It is a good idea to start by etching or cutting a very narrow line in the foil. If the detector is too sensitive, the line can be widened.
When the moisture sensor is dry, it has a very high resistance. As it becomes moist, its resistance decreases, but it's often on the order of several hundred thousand ohms.
A simple detector circuit that can be used with a moisture detector is shown in ill. 2-30. Here two npn transistors are connected in a Darlington arrangement that will provide a very high current gain. Even a very small current through the moisture detector will cause base current to flow in transistor Q1. This in turn will cause even more base current to flow in transistor Q2. The result is that a very small amount of current through the moisture detector will cause the relay in the emitter circuit to operate.
Updated: Wednesday, March 23, 2011 19:29 PST