Instrumentation Sensors Book
.pdf
5.5 Basic Processor |
75 |
Digital data inputs
Enable address
Data output
Address decoder
Figure 5.20 Digital multiplexer.
5.4.3Programmable Logic Arrays
Many systems have large blocks of gates to perform custom logic and sequential logic functions. These functions were constructed using the SN 74 family of logic gates. The logic gates can now be replaced with a programmable logic array (PLA). One of these devices replaces many gate devices, requires less power, and can be configured (programmed) by the end user to perform all of the required system functions. The devices also have the flexibility to be reprogrammed if an error in the logic is found or there is a need to upgrade the logic. Because they can be programmed by the end user, there is no wasted time, unlike with the Read Only Memory (ROM), which had to be programmed by the manufacturer. However, the Electrically Erasable Programmable ROM (EEPROM) is available as a reprogrammable device, and technology-wise is similar to the PLA.
5.4.4Other Interface Devices
A number of controller peripheral devices, such as timing circuits, are commercially available. These circuits are synchronized by clock signals from the controller that are referenced to very accurate crystal oscillators, which are accurate to within less than 0.001%. Using counters and dividers, the clock signal can be used to generate very accurate delays and timing signals. Compared to RC-generated delays and timing signals, which can have tolerances of more than 10%, the delays and timing signals generated by digital circuits in new equipment are preferred.
Other peripheral devices are digital comparators, encoders, decoders, display drivers, counters, shift registers, and serial to parallel converters.
5.5Basic Processor
Figure 5.21 shows a simplified diagram of a digital processor. The system typically consists of the central processing unit (CPU) with arithmetic unit (ALU), random access memory (RAM), read only memory (ROM), and data input/output ports. Communication between the units uses three buses: (1) a two-way data bus that
76 |
|
|
|
Digital Electronics |
|
Processor |
Two way data bus |
|
|
||
|
Central |
|
|
|
|
Input |
processing |
Program |
Data |
Output |
|
CPU control |
memory |
memory |
|||
ports |
ports |
||||
and arithmetic |
ROM |
RAM |
|||
|
|
||||
|
unit |
|
|
|
|
Control bus |
Address bus |
External memory, ADC, DAC, multiplexer, etc
Figure 5.21 Digital processor block diagram.
passes data between all of the individual units; (2) a one-way address bus from the CPU that gives the address to which the data is to be sent or retrieved; and (3) a one-way control bus that selects the unit to which the data and address is to be sent, or the address from which data is to be retrieved. The input/output ports are used for communication with other computers or peripheral units. The processor forms the heart of the controller, as shown in Chapter 14, Figure 14.3. The input port can receive analog sensor data via ADCs and multiplexers. Once received, the data is first conditioned by stored equations or lookup tables to correct for linearity, offset, span, or temperature. The new data is then compared to the preset value, and the appropriate signal sent via the output port and a DAC to control an actuator. The input port can receive serial or parallel formatted data. In the case of serial data, it is converted to parallel data by a serial to parallel converter for internal use. The controller has the ability to serially scan a large number of sensors via the input port, and send data to a number of actuators via the output port. The sensors are scanned every few milliseconds, and the output data to the actuators updated, giving continuous monitoring and control. The controller can be used for sequential control, as well as continuous monitoring and control.
5.6Summary
Digital electronics were introduced in this chapter not only as a refresher, but also to extend digital concepts to their applications in process control. Physical variables are analog and the central processor is digital; therefore, various methods of converting analog to digital and digital to analog were discussed. Analog data can be converted to digital using successive approximation for low-speed applications, or by using flash conversion for high speed, when converting digital to analog weighted resistor techniques for an analog voltage output, or pulse width modulation for power control. Data acquisition devices are used to feed data to the central
5.6 Summary |
77 |
processor. The use of analog and digital multiplexers in data acquisition systems is shown with demultiplexers, and in the basic processor block diagram.
The use of comparators in analog to digital and digital to analog converters was discussed. The various methods of conversion and their relative merits were given, along with a discussion on analog to frequency converters.
References
[1]Jones, C. T., Programmable Logic Controllers, 1st ed., Patrick-Turner Publishing Co., 1996, pp. 17–22.
[2]Tokheim, R. L., Digital Electronics Principles and Applications, 6th ed., Glencoe/ McGraw-Hill, 2003, pp. 77–113.
[3]Humphries, J. T., and L. P. Sheets, Industrial Electronics, 4th ed., Delmar, 1993, pp. 566–568.
[4]Miller, M. A., Digital Devices and Systems with PLD Applications, 1st ed., Delmar, 1997,
pp.452–486.
[5]McGonigal, J., “Integrated Solutions for Calibrated Sensor Signal Conditioning,” Sensor Magazine, Vol. 20, No. 10, September 2003.
[6]Madni, A., J. B. Vuong, and P. T. Vuong, “A Digital Programmable Pulse-Width-Modula- tion Converter,” Sensors Magazine, Vol. 20, No. 5, May 2003.
[7]Banerjee, B., “The 1 Msps Successive Approximation Register A/D Converter,” Sensors Magazine, Vol. 18, No. 12, December 2001.
[8]Contadini, F., “Demystifying Sigma-Delta ADCs,” Sensors Magazine, Vol. 19, No. 8, August 2002.
[9]Khalid, M., “Working at High Speed: Multi-megahertz 16-Bit A/D Conversion,” Sensors Magazine, Vol. 15, No. 5, May 1998.
[10]Johnson, C. D., Process Control Instrumentation Technology, 7th ed., Prentice Hall, 2003,
pp.147–158.
C H A P T E R 6
Microelectromechanical Devices and
Smart Sensors
6.1Introduction
The development of new devices in the microelectronics industry has over the past 50 years been responsible for producing major changes in all industries. The technology developed has given cost-effective solutions and major improvements in all areas. The microprocessor is now a household word, and is embedded in every appliance, entertainment equipment, most toys, and every computer in the home. In the process control industry, processes and process control have been refined to a level only dreamed of a few years ago. The major new component is the microprocessor, but other innovations in the semiconductor industry have produced accurate sensors for measuring temperature, time, and light intensity; along with microelectromechanical devices for measuring pressure, acceleration, and vibration. This has certainly brought about a major revolution in the process control industry. Silicon has been the semiconductor of choice. Silicon devices have a good operating range (−50° to +150°C), a low leakage, and can be mass-produced with tight tolerances. The processing of silicon has been refined, so that many millions of devices can be integrated and produced on a single chip, which enables a complete electronic system to be made in a package. Several unique properties of silicon make it a good material for use in sensing physical parameters. Some of these properties are as follows:
•The piezoresistive effect can be used in silicon for making strain gauges;
•The Hall Effect or transistor structures can be used to measure magnetic field strength;
•Linear parametric variation with temperature makes it suitable for temperature measurement;
•Silicon has light-sensitive parameters, making it suitable for light intensity measurements;
•Silicon does not exhibit fatigue, and has high strength and low density, making it suitable for micromechanical devices.
Micromechanical sensing devices can be produced as an extension of the standard silicon device process, enabling the following types of sensors:
79
80 |
Microelectromechanical Devices and Smart Sensors |
•Pressure;
•Force and strain;
•Acceleration;
•Vibration;
•Flow;
•Angular rate sensing;
•Frequency filters.
Micromechanical devices are very small, have low mass, and normally can be subjected to high overloads without damage.
6.2Basic Sensors
Certain properties of the semiconductor silicon crystal can be used to sense physical properties, including temperature, light intensity, force, and magnetic field strength.
6.2.1Temperature Sensing
A number of semiconductor parameters vary linearly with temperature, and can be used for temperature sensing. These parameters are diode or transistor junction voltages, zener diode voltages, or polysilicon resistors. In an integrated circuit, the fact that the differential base emitter voltage between two transistors operating at different current densities (bandgap) is directly proportional to temperature is normally used to measure temperature. The output is normally adjusted to give a sensitivity of 10 mV per degree (C, F, or K). These devices can operate over a very wide supply voltage range, and are accurate to ±1°C or K and ±2°F, over the temperature range −55° to +150°C [1]. The device families manufactured by National Semiconductor are as follows:
•LM35, which gives 10 mV/°C;
•LM34, which gives 10 mV/°F;
•LM135, which gives 10 mV/K (the output voltage = 2.73V at 0°C).
A simplified block diagram of the LM75 is shown in Figure 6.1. This device is calibrated in Celsius, has a 10-bit DAC, and a register for digital readout. The analog voltage signal from the temperature sensor is digitized in a 9-bit Delta-Sigma ADC with a 10-bit decimation filter. Three address pins (A0, A1, A2) are available, so that any one of eight devices can be selected when connected to a common bus. An overtemperature alarm pin is available that can be programmed from the two-way data bus [2].
6.2.2Light Intensity
Semiconductor devices are in common use as photointensity sensors. Photodiodes, phototransistors, and integrated photosensors are commercially available [3]. Integrated devices have on-chip temperature compensation and high sensitivity, and can
6.2 Basic Sensors |
81 |
+ Ve supply
|
Silicon |
9 Bit A/D |
Over |
|
|
bandgap |
|
||
|
converter |
temperature |
|
|
|
temperature |
|
||
|
10 Bit Filter |
protection |
Alarm |
|
|
sensor |
|||
|
|
|
output |
|
|
|
|
|
|
A0 |
Control |
|
|
Serial data |
A1 |
|
Interface |
||
logic |
|
enable |
||
A2 |
|
|
||
|
|
|
|
Figure 6.1 LM75 Block diagram.
be configured to have a voltage, digital, or frequency output that is proportional to intensity. The devices also can be made sensitive to visual or infrared frequency spectra. Photons cause junction leakage in photodiodes that is proportional to light intensity; thus, the reverse leakage is a measure of light intensity. In photodiodes, the effect is to increase the base current of the device giving an amplified collec- tor-emitter current. Devices are available with sensitivities of 80 mV per 
W/cm2 (880 nm) and a linearity of 0.2%. Programmable light to frequency converters can sense light intensities of 0.001 to 100 k
W/cm2 with low temperature drift, and an absolute frequency tolerance of ±5%. Some typical devices families manufactured by Texas Instrument are as follows:
•TSL230 Light to frequency converter;
•TSL235 Light to frequency converter;
•TSL250 Light to voltage converter;
•TSL260 IR Light to voltage converter.
Figure 6.2 shows the circuit of the front end of an integrated temperature-com- pensated phototransistor. The circuit is similar to that of the front end of an op-amp. In the op-amp, the bases of the differential input pair of transistors are driven from the differential input signals. In this case, the bases are joined together and internally biased. One of the input transistors is typically screened with metallization, while the other is not screened, so that incident light from a window in the package can fall on the transistor, which will then become a phototransistor. The light intensity is converted into an electrical signal and amplified. Because the input stage is a differential stage, it is temperature-compensated. The current supply is typically from a bandgap regulator [4].
6.2.3Strain Gauges
In a resistive type of strain gauge, the gauge factor (GF) is the fractional change in resistance divided by the fractional change in length, and is given by:
82 |
Microelectromechanical Devices and Smart Sensors |
+ Ve |
|
Supply |
|
|
Output to |
|
next stage |
Phototransistor |
|
with window |
|
Set |
|
zero |
Current |
|
|
|
supply |
−Ve
−Supply
Figure 6.2 Temperature-compensated integrated phototransistor circuit.
GF = ∆R/R/∆l/l = ∆R/R/strain |
(6.1) |
where ∆R/R is the fractional change in resistance, and ∆l/l is the fractional change in length or strain.
The semiconductor strain gauge is more sensitive than the deposited resistive strain gauge, since it uses the piezoresistive effect, which can be very large. The physical dimension change then can be ignored. The gauge factor can be either positive or negative, depending upon doping, and can be 100 times higher than for foil gauges. The gauge factor in silicon depends on the crystal orientation, and the type of doping of the gauge, which can be n or p. Strain gauges are normally p-doped resistors, since these have the highest gauge factor. Lightly doped or higher valued resistors have higher gauge factors, but are more temperature-sensitive. Thus, the resistance is a compromise between gauge factor and temperature sensitivity. The semiconductor gauge is very small (0.5 × 0.25 mm), does not suffer from fatigue, and is commercially available with a compensating device perpendicular to the measuring device, or with four elements to form the arms of a bridge. The strain gauge also can be integrated with compensating electronics and amplifiers to give conditioning and high sensitivity [5].
6.2.4Magnetic Field Sensors
Magnetic fields can be sensed using the Hall Effect, magnetoresistive elements (MRE), or magnetotransistors [6]. Some applications for magnetic field sensors are given in Section 11.2.1.
The Hall Effect occurs when a current flowing in a carrier (semiconductor) experiences a magnetic field perpendicular to the direction of the current flow. The interaction between the two causes the current to be deflected perpendicular to the
6.2 Basic Sensors |
83 |
direction of both the magnetic field and the current. Figure 6.3 shows the effect of a magnetic field on the current flow in a Hall Effect device. Figure 6.3(a) shows the current flow without a magnetic field, and Figure 6.3(b) shows the deflection of the current flow with a magnetic field, which produces the Hall voltage. Table 6.1 gives the characteristics of some common materials used as Hall Effect devices.
The MRE is the property of a current-carrying ferromagnetic material to change its resistance in the presence of a magnetic field. As an example, a ferromagnetic (permalloy) element (20% iron and 80% nickel) will change its resistivity approximately 3% when the magnetic field is rotated 90°. The resistivity rises to a maximum when the current and magnetic field are coincident to each other, and are at a minimum when the magnetic field and current are perpendicular to each other. The effect of rotating an MRE is shown in Figure 6.4(a). The attribute is known as the anisotropic magnetoresistive effect. The resistance R, or an element, is related to the angle q between the current and the magnetic field directions by the expression:
R = R11 cos2 q + R1 sin2 q |
(6.2) |
where R11 is the resistance when the current and magnetic fields are parallel, and R1 is the resistance when the current and magnetic fields are perpendicular.
MRE devices give an output when stationary, which makes them suitable for zero speed sensing, or position sensing. For good sensitivity and to minimize temperature effects, four devices are normally arranged in a Wheatstone bridge configuration. In an MRE device, aluminum strips are put 45° across the permalloy element to linearize the device, as shown in Figure 6.4(b). The low resistance aluminum strips cause the current to flow 45° to the element, which biases the element into a linear operating region. Integrated MRE devices can typically operate from −40 to +150°C at frequencies up to 1 MHz.
|
No magnetic field |
Magnetic field |
|
|
|
Current flow |
Current flow |
Current flow |
|
||
|
v |
v |
Semiconductor plate |
|
Semiconductor plate |
|
Hall voltage |
Hall voltage |
|
(a) |
(b) |
Figure 6.3 Hall Effect device (a) without a magnetic field, and (b) with a magnetic field.
Table 6.1 Hall Effect Sensitivities
Material |
Temperature Range Supply Voltage Sensitivity @ 1 kA/m Frequency Range |
Indium −40° to +100°C GaAs −40° to +150°C Silicon −40° to +150°C
1V |
7 mV |
0 to 1 MHz |
5V |
1.2 mV |
0 to 1 MHz |
12V |
94 mV |
0 to 100 kHz |
84 |
Microelectromechanical Devices and Smart Sensors |
Magnetic flux |
Magnetic flux |
|
Current |
|
Element |
|
direction |
Resistivity
|
|
|
|
Permalloy |
|
|
|
|
element |
0 |
45 |
90 |
135 |
180 |
|
Angle (degrees) |
|
Aluminum |
|
|
|
stripes |
||
|
|
|
|
|
|
|
(a) |
|
(b) |
Figure 6.4 MRE devices: (a) effect of direction of magnetic field on element, and (b) element layout.
Magnetotransistors can be made using bipolar or CMOS technology. Figure 6.5(a) shows the topology of a PNP bipolar magnetotransistor. The electron flow without a magnetic field is shown in Figure 6.5(a), and the electron flow in the presence of a magnetic field is shown in Figure 6.5(b), and the junction cross section in shown Figure 6.5(c). The device has two collectors as shown. When the base is forward biased, the current from the emitter is equally divided between the two collectors when no magnetic field is present. When a magnetic field is present, the current flow is deflected towards one of the collectors, similar to the Hall Effect device. This gives an imbalance between the current in the two collectors, which is proportional to the magnetic field strength, can be amplified as a differential signal, and can be used to measure the strength of the magnetic field [7]. A comparison of the sensitivities of magnetic field sensors is given in Table 6.2.
6.3Piezoelectric Devices
The piezoelectric effect is the coupling between the electrical and mechanical properties of certain materials. If a potential is applied across piezoelectric material, then a mechanical change occurs. This is due to the nonuniform charge distribution
Table 6.2 Comparison of the Sensitivities of Magnetic Field Sensors
Sensor Type |
Temperature Range |
Sensitivity @ 1 kA/m |
Frequency Range |
Mechanical Stress |
Hall Effect (Si) |
−40° to +150°C |
90 mV |
0 to 100 kHz |
High |
MRE |
−40° to +150°C |
140 mV |
0 to 1 MHz |
Low |
Magnetotransistor |
−40° to +150°C |
250 mV |
0 to 500 kHz |
Low |