Module 2

Measurement Systems
Version 2 EE IIT, Kharagpur 1

Lesson 6
Displacement and Speed Measurement
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
At the end of this lesson, the student should be able to • • • • • • Name three methods of displacement measurement using passive electrical sensors. Sketch the construction and characteristics of LVDT. Explain the principles of operation of inductive and capacitive types of proximity sensors. Distinguish between variable distance and variable area type of capacitance displacement sensors. Sketch and explain the principle of operation of a optical type displacement sensor. Name two methods of noncontact type speed sensing and explain their principles of operation.

Introduction
Displacement and speed are two important parameters whose measurements are important in many position and speed control schemes. Error free measurements of these two parameters are necessary in order to achieve good control performance. Displacement measurement can be of different types. The displacement may be in the range of few μm to few cm. Moreover the measurement may be of contact type or noncontact type. Again displacement to be measured can be linear or angular (rotary). Similar is the case for speed measurement. Accordingly different measuring schemes are used for measurement of these two parameters. In this lesson, we shall discuss about few such schemes.

Displacement Measurement
Broadly speaking, displacement measurement can be of two types: contact and noncontact types. Besides the measurement principles can be classified into two categories: electrical sensing and optical sensing. In electrical sensing, passive electrical sensors are used variation of either inductance or capacitance with displacement is measured. On the other hand the optical method mainly works on the principle of intensity variation of light with distance. Interferometric technique is also used for measurement of very small displacement in order of nanometers. But this technique is more suitable for laboratory purpose, not very useful for industrial applications.

Potentiometer
Potentiometers are simplest type of displacement sensors. They can be used for linear as well as angular displacement measurement, as shown in Fig. 1. They are the resistive type of transducers and the output voltage is proportional to the displacement and is given by: x eo = i E , xt where xi is the input displacement, xt is the total displacement and E is the supply voltage.

Version 2 EE IIT, Kharagpur 3

The major problem with potentiometers is the contact problem resulting out of wear and tear between the moving and the fixed parts. As a result, though simple, application of potentiometers is limited.

θi E xt xi eo eo (a) Fig. 1 Potentiometer (a) Linear (b) Rotary E

(b)

Linear Variable Differential transformer (LVDT)
LVDT works on the principle of variation of mutual inductance. It is one of the most popular types of displacement sensor. It has good linearity over a wide range of displacement. Moreover the mass of the moving body is small, and the moving body does not make any contact with the static part, thus minimizing the frictional resistance. Commercial LVDTs are available with full scale displacement range of ± 0.25mm to ± 25mm. Due to the low inertia of the core, the LVDT has a good dynamic characteristics and can be used for time varying displacement measurement range. The construction and principle of operation of LVDT can be explained with Fig. 2(a) and Fig. 2(b). It works on the principle of variation of the mutual inductance between two coils with displacement. It consists of a primary winding and two identical secondary windings of a transformer, wound over a tubular former, and a ferromagnetic core of annealed nickel-iron alloy moves through the former. The two secondary windings are connected in series opposition, so that the net output voltage is the difference between the two. The primary winding is excited by 1-10V r.m.s. A.C. voltage source, the frequency of excitation may be anywhere in the range of 50 Hz to 50 KHz. The output voltage is zero when the core is at central position (voltage induced in both the secondary windings are same, so the difference is zero), but increasing as the core moves away from the central position, in either direction. Thus, from the measurement of the output voltage only, one cannot predict, the direction of the core movement. A phase sensitive detector (PSD) is a useful circuit to make the measurement direction sensitive. It is connected at the output of the LVDT and compares the phase of the secondary output with the primary signal to judge the direction of movement. The output of the phase sensitive detector after low pass filtering becomes a d.c voltage for a steady deflection. The output voltage after PSD vs. displacement characteristics is shown in Fig. 2(c).

Version 2 EE IIT, Kharagpur 4

V0 (Measurement)

Primary Coil Insulating Form

Core Displacement X (Measurand) Ferromagnetic Core

Secondary Coil

Vref

Secondary Coil

Fig. 2(a) Construction of LVDT.

vo

Secondary Coil

Core

Primary Coil vref Fig. 2(b) Series opposition connection of secondary windings.

Version 2 EE IIT, Kharagpur 5

vo (d.c)
Voltage Level

Displacement x

Linear Range

Fig. 2(c) Output voltage vs. displacement characteristics of LVDT after Phase sensitive detection.

Inductive type Sensors
LVDT works on the principle of variation of mutual inductance. There are inductive sensors for measurement of displacement those are based on the principle of variation of self inductance. These sensors can be used for proximity detection also. Such a typical scheme is shown in Fig. 3. In this case the inductance of a coil changes as a ferromagnetic object moves close to the magnetic former, thus change the reluctance of the magnetic path. The measuring circuit is usually an a.c. bridge. AC Supply vref Inductance Measuring Circuit

Ferromagnetic Target Object

X (Measurand)

Fig. 3 Schematic diagram of a self inductance type proximity sensor. Version 2 EE IIT, Kharagpur 6

Rotary Variable Differential Transformer (RVDT)
Its construction is similar to that of LVDT, except the core is designed in such a way that when it rotates the mutual inductance between the primary and each of the secondary coils changes linearly with the angular displacement. Schematic diagram of a typical RVDT is shown in Fig. 4. vo

Secondary Coil Core

Primary Coil

vref θ

Fig. 4 Rotary Variable Differential Transformer (RVDT)

Resolver
Resolvers also work on the principle of mutual inductance variation and are widely used for measurement of rotary motion. The basic construction is shown in Fig. 5. A resolver consists of a rotor containing a primary coil and two stator windings (with equal number of turns) placed perpendicular to each other. The rotor is directly attached to the object whose rotation is being measured. If a.c. excitation of r.m.s voltage Vr is applied, then the induced voltages at two stator coils are given by: v01 = KVr cosθ and

v02 = KVr sin θ ; where K is a constant.

By measuring these two voltages the angular position can be uniquely determined, provided (0 ≤ θ ≤ 900 ) . Phase sensitive detection is needed if we want to measure for angles in all the four quadrants. Synchros work widely as error detectors in position control systems. The principle of operation of synchros is similar to that of resolvers. However it will not be discussed in the present lesson.

Version 2 EE IIT, Kharagpur 7

Output vo2 = Kvref sin θ

Rotor

AC Supply vref

θ Stator Output vo1 = Kvref cos θ

Stator

Fig. 5 Schematic diagram of the resolver.

Capacitance Sensors
The capacitance type sensor is a versatile one; it is available in different size and shape. It can also measure very small displacement in micrometer range. Often the whole sensor is fabricated in a silicon base and is integrated with the processing circuit to form a small chip. The basic principle of a capacitance sensor is well known. But to understand the various modes of operation, consider the capacitance formed by two parallel plates separated by a dielectric. The capacitance between the plates is given by: ε ε A (1) C= r 0 d where A=Area of the plates d= separation between the plates ε r = relative permittivity of the dielectric

ε 0 = absolute permittivity in free space = 8.854 × 10 −12 F / m .
A capacitance sensor can be formed by either varying (i) the separation (d), or, (ii) the area (A), or (iii) the permittivity ( ε r ). A displacement type sensor is normally based on the first two (variable distance and variable area) principles, while the variable permittivity principle is used for measurement of humidity, level, etc. Fig.6 Shows the basic constructions of variable gap and variable area types of capacitance sensors mentioned above. Fig. 6(a) shows a variable distance type sensor, where the gap between the fixed and moving plates changes. On the other hand, the area of overlap between the fixed plate and moving plate changes in Fig. 6(b), maintaining the gap constant. The variable area type sensor gives rise to linear variations of capacitance with the input variable, while a variable separation type sensor follows inverse relationship. Version 2 EE IIT, Kharagpur 8

Fixed

(a)

(b)

Fig.6 Capacitive type displacement Sensors: a) variable separation type, b) variable area type. Capacitance sensors are also used for proximity detection. Such a typical scheme is shown in Fig. 4. Capacitive proximity detectors are small in size, noncontact type and can detect presence of metallic or insulating objects in the range of approximately 0-5cm. For detection of insulating objects, the dielectric constant of the insulating object should be much larger than unity. Fig. 7 shows the construction of a proximity detector. Its measuring head consists of two electrodes, one circular (B) and the other an annular shaped one (A); separated by a small dielectrical spacing. When the target comes in the closed vicinity of the sensor head, the capacitance between the plates A and B would change, which can be measured by comparing with a fixed reference capacitor. The measuring circuits for capacitance sensors are normally capacitive bridge type. But it should be noted that, the variation of capacitance in a capacitance type sensor is generally very small (few pF only, it can be even less than a pF in certain cases). These small changes in capacitor, in presence of large stray capacitance existing in different parts of

Target

x

A

B

Fig. 7 Capacitance Proximity Detector.

Version 2 EE IIT, Kharagpur 9

the circuit are difficult. So the output voltage would generally be noisy, unless the sensor is designed and shielded carefully, the measuring circuit should also be capable of reducing the effects of stray fields.

Optical Sensors
Optical displacement sensors work on the basic principle that the intensity of light decreases with distance. So if the source and detector are fixed, the amount of light reflected from a moving surface will depend on the distance of the moving surface from the fixed ones. Measurement using this principle requires proper calibration since the amount of light received depends upon the reflectivity of the surface, intensity of the source etc. Yet it can provide a simple method for displacement measurement. Optical fibers are often used to transmit light to and from the measuring zone. Such a scheme with bundle fibers is shown in Fig. 8. It uses two bundle fibers, one for transmitting light from the source and the other to the detector. Light reflected on the receiving fiber bundle by the surface of the target object is carried to a photo detector. The light source could be Laser or LED; photodiodes or phototransistors are used for detection.
Laser, LED, etc. Power Source Light Source Photodiode, Phototransistor, etc. Signal Processor Photodetector Receiving X Fibers (Measurand) Transmitting Fibers

Target Object

Position Measurement

Fig. 8 A Fiber optic position sensor.

Speed Measurement
The simplest way for speed measurement of a rotating body is to mount a tachogenerator on the shaft and measure the voltage generated by it that is proportional to the speed. However this is a contact type measurement. There are other methods also for noncontact type measurements. The first method is an optical method shown in Fig. 9. An opaque disc with perforations or transparent windows at regular interval is mounted on the shaft whose speed is to be measured. A LED source is aligned on one side of the disc in such a way that its light can pass through the transparent windows of the disc. As the disc rotates the light will alternately passed through the transparent windows and blocked by the opaque sections. A photodetector fixed on the other side of the disc detects the variation of light and the output of the detector after signal conditioning would be a square wave (as shown) whose frequency is decided by the speed and the number of holes (transparent windows) on the disc.

Version 2 EE IIT, Kharagpur 10

vo
Output

t
LED Light Source

Transparent Windows Photocell, Phototransistor, or Photodiode (Light Sensor) Opaque Disk

Fig. 9 Schematic arrangement of optical speed sensing arrangement. Fig.10 shows another scheme for speed measurement. It is a variable reluctance type speed sensor. A wheel with projected teethes made of a ferromagnetic material is mounted on the shaft whose speed is to be measured. The static sensor consists of a permanent magnet and a search coil mounted on the same assembly and fixed at a closed distance from the wheel. The flux through the permanent magnet completes the path through the teeth of the wheel and cut the search coil. As the wheel rotates there would be change in flux cut and a voltage will be induced in the search coil. The variation of the flux can be expressed as: φ (t ) = φo + φm sin ω t (2) where ω is the angular speed of the wheel. Then the voltage induced in the coil is:

e = −N

dφ = − Nω φm cos ω t dt

(3)

where N is the number of turns in the search coil. So both the amplitude and frequency of the induced voltage is dependent on the speed of rotation. This voltage is fed to a comparator circuit that gives a square wave type voltage whose amplitude is constant, but frequency is proportional to the speed. A frequency counter is used to count the number of square pulses during a fixed interval and displays the speed.
d Yoke Permanent magnet

Search Coil Terminals Coil

Fig. 10 Variable reluctance type speed sensor. Version 2 EE IIT, Kharagpur 11

Conclusion
Few techniques commonly used for displacement and speed measurements have been discussed in this lesson. The selection of the sensing scheme depends on the requirement, environment and accessibility. Displacement/position sensing can be done in two ways. One method is to convert the displacement signal into variation of inductance or capacitance and then use suitable measuring circuit to measure their variation. On the other hand, in optical method the intensity of the light reflected from a moving surface is measured and calibrated in terms of the distance. An important application of displacement measurement is proximity sensing. Few such schemes have been discussed in this lesson. However eddy current type proximity sensing has not been discussed in this lesson. The most popular type of speed sensor is the tachogenerator. The tachogenerator is mounted on the shaft and the voltage induced that is proportional to the speed is measured. But there are other noncontact methods also in which the speed signal is converted into frequency signal and the frequency is measured. Two such techniques have been discussed in this lesson.

References
1. J.P. Bentley: Principles of Measurement Systems (3/e), Longman, U.K., 1995. 2. E.O. Doeblin: Measurement System Application and Design (4/e), Mcgraw-Hill, Singapore, 1990. 3. L.K.Baxter: Capacitance Sensors Design and Applications, IEEE Press, New Jersey, 1997. 4. D. Patranabis: Sensors and Transducers (2/e), PHI, New Delhi, 2003. 5. C.W. de Silva: Control Sensors and Actuators, Prentice Hall, New Jersey, 1989.

Review Questions
1. What is the function of a Phase Sensitive Detector (PSD) in a LVDT circuit? 2. Discuss the construction and operating principle of a LVDT. 3. Distinguish between variable gap and variable area type capacitance displacement sensors. 4. What are the advantages and limitations of optical displacement (position) sensors? 5. Name two method of noncontact type speed measurement. Explain with a schematic diagram the principle of operation of any one of them. 6. An optical type speed sensor has a disc with 36 rectangular holes placed at regular intervals on the periphery of the disc. The frequency of the photodetector output is 360 Hz. Find the speed of the shaft in rpm on which the disc is mounted.

Answer
6. 600 rpm.

Version 2 EE IIT, Kharagpur 12

Module 1
Introduction
Version 2 EE IIT, Kharagpur 1

Lesson 1
Introduction to Industrial Automation and Control
Version 2 EE IIT, Kharagpur 2

Lesson Objectives
• • • • • To define Automation and Control and explain the differences in the sense of the terms To explain the relation between Automation and Information Technology To underline the basic objectives of a manufacturing industry and explain how automation and control technologies relate to these To introduce the concept of a Product Life Cycle and explain how Automation and Control technologies relate to the various phases of the cycle To classify Manufacturing plants and categorise the different classes of Automation Systems that are appropriate for these

Understanding the Title of the Course
Let us first define the three key words in the title, namely,

Industry
In a general sense the term “Industry” is defined as follows. Definition: Systematic Economic Activity that could be related to

Manufacture/Service/ Trade. In this course, we shall be concerned with Manufacturing Industries only.

Automation
The word ‘Automation’ is derived from greek words “Auto”(self) and “Matos” (moving). Automation therefore is the mechanism for systems that “move by itself”. However, apart from this original sense of the word, automated systems also achieve significantly superior performance than what is possible with manual systems, in terms of power, precision and speed of operation. Definition: Automation is a set of technologies that results in operation of machines and systems without significant human intervention and achieves performance superior to manual operation

A Definition from Encyclopaedia Britannica
The application of machines to tasks once performed by human beings or, increasingly, to tasks that would otherwise be impossible. Although the term mechanization is often used to refer to the simple replacement of human labour by machines, automation generally implies the integration of machines into a selfgoverning system.

Version 2 EE IIT, Kharagpur 3

Point to Ponder: 1
A. Why does an automated system achieve superior performance compared to a manual one? B. Can you give an example where this happens?

Control
It is perhaps correct to expect that the learner for this course has already been exposed to a course on Control Systems, which is typically introduced in the final or pre-final year of an undergraduate course in Engineering in India. The word control is therefore expected to be familiar and defined as under. Definition: Control is a set of technologies that achieves desired patterns of variations of operational parameters and sequences for machines and systems by providing the input signals necessary.

Point to Ponder: 2
A. Can you explain the above definition in the context of a common control system, such as temperature control in an oven? B. Is the definition applicable to open-loop as well as closed loop control? It is important at this stage to understand some of the differences in the senses that these two terms are generally interpreted in technical contexts and specifically in this course. These are given below. 1. Automation Systems may include Control Systems but the reverse is not true. Control Systems may be parts of Automation Systems. 2. The main function of control systems is to ensure that outputs follow the set points. However, Automation Systems may have much more functionality, such as computing set points for control systems, monitoring system performance, plant startup or shutdown, job and equipment scheduling etc. Automation Systems are essential for most modern industries. It is therefore important to understand why they are so, before we study these in detail in this course.

Point to Ponder: 3
A. Can you give an example of an automated system, which contains a control system as a part of it? B. What are the other parts of the system?

Version 2 EE IIT, Kharagpur 4

Industrial Automation vs. Industrial Information Technology
Industrial Automation makes extensive use of Information Technology. Fig. 1.1 below shows some of the major IT areas that are used in the context of Industrial Automation.

Industrial IT
Control and Signal Processing Simulation, Design, Analysis, Optimization Communication and Networking Real-time Computing

Database

Fig. 1.1 Major areas of IT which are used in the context of Industrial Automation.

Point to Ponder: 4
A. Try to find an example automated system which uses at least one of the areas of Industrial IT mentioned in Fig. 1.1 (Hint: Try using the internet) However, Industrial Automation is distinct from IT in the following senses A. Industrial Automation also involves significant amount of hardware technologies, related to Instrumentation and Sensing, Actuation and Drives, Electronics for Signal Conditioning, Communication and Display, Embedded as well as Stand-alone Computing Systems etc. B. As Industrial Automation systems grow more sophisticated in terms of the knowledge and algorithms they use, as they encompass larger areas of operation comprising several units or the whole of a factory, or even several of them, and as they integrate manufacturing with other areas of business, such as, sales and customer care, finance and the entire supply chain of the business, the usage of IT increases dramatically. However, the lower level Automation Systems that only deal with individual or , at best, a group of machines, make less use of IT and more of hardware, electronics and embedded computing.

Point to Ponder: 5
A. Can you give an example of an automated system, some of whose parts makes a significant application of Industrial IT?

Version 2 EE IIT, Kharagpur 5

B. Can you give an example of an automated system, none of whose parts makes a significant application of Industrial IT? Apart from the above, there are some other distinguishing features of IT for the factory that differentiate it with its more ubiquitous counterparts that are used in offices and other business. A. Industrial information systems are generally reactive in the sense that they receive stimuli from their universe of discourse and in turn produce responses that stimulate its environment. Naturally, a crucial component of an industrial information system is its interface to the world. B. Most of industrial information systems have to be real-time. By that we mean that the computation not only has to be correct, but also must be produced in time. An accurate result, which is not timely may be less preferable than a less accurate result produced in time. Therefore systems have to be designed with explicit considerations of meeting computing time deadlines. C. Many industrial information systems are considered mission-critical, in the sense that the malfunctioning can bring about catastrophic consequences in terms of loss of human life or property. Therefore extraordinary care must be exercised during their design to make them flawless. In spite of that, elaborate mechanisms are often deployed to ensure that any unforeseen circumstances can also be handled in a predictable manner. Fault-tolerance to emergencies due to hardware and software faults must often be built in.

Point to Ponder: 6
A. Can you give an example of an automated system, which is reactive in the sense mentioned above? B. Can you give an example of an automated system, which is real-time in the sense mentioned above C. Can you give an example of an automated system, which is mission-critical in the sense mentioned above

Role of automation in industry
Manufacturing processes, basically, produce finished product from raw/unfinished material using energy, manpower and equipment and infrastructure. Since an industry is essentially a “systematic economic activity”, the fundamental objective of any industry is to make profit. Roughly speaking, Profit = (Price/unit – Cost/unit) x Production Volume (1) So profit can be maximised by producing good quality products, which may sell at higher price, in larger volumes with less production cost and time. Fig 1.2 shows the major parameters that affect the cost/unt of a mass-manufactured industrial product.

Version 2 EE IIT, Kharagpur 6

Cost/unit

Material

Energy

Manpower

Infrastructure

Fig. 1.2 The Components of per unit Manufacturing Cost Automation can achieve all these in the following ways, Figure 1.4 shows how overall production time for a product is affected by various factors. Automation affects all of these factors. Firstly, automated machines have significantly lower production times. For example, in machine tools, manufacturing a variety of parts, significant setup times are needed for setting the operational configuration and parameters whenever a new part is loaded into the machine. This can lead to significant unproductive for expensive machines when a variety of products is manufactured. In Computer Numerically Controlled (CNC) Machining Centers set up time is reduced significantly with the help of Automated Tool Changers, Automatic Control of Machines from a Part Program loaded in the machine computer. Such a machine is shown in Figure 1.3. The consequent increase in actual metal cutting time results in reduced capital cost and an increased volume of production.

Point to Ponder: 7
A. With reference to Eq. (1), explain how the following automation systems improve industrial profitability. a. Automated Welding Robots for Cars b. Automated PCB Assembly Machines c. Distributed Control Systems for Petroleum Refineries

Fig. 1.3 A CNC Machine with an Automated Tool Changer and the Operator Console with Display for Programming and Control of the Machine Version 2 EE IIT, Kharagpur 7

Production Volume

Production Time

Material Handling Time

Idle Time

Quality Assurance Time

Fig. 1.4 The major factors that contribute to Overall Production Time Similarly, systems such as Automated Guided Vehicles, Industrial Robots, Automated Crane and Conveyor Systems reduce material handling time. Automation also reduces cost of production significantly by efficient usage of energy, manpower and material. The product quality that can be achieved with automated precision machines and processes cannot be achieved with manual operations. Moreover, since operation is automated, the same quality would be achieved for thousands of parts with little variation. Industrial Products go through their life cycles, which consists of various stages. At first, a product is conceived based on Market feedbacks, as well as Research and Development Activities. Once conceived the product is designed. Prototype Manufacturing is generally needed to prove the design. Once the design is proved, Production Planning and Installation must be carried out to ensure that the necessary resources and strategies for mass manufacturing are in place. This is followed by the actual manufacture and quality control activities through which the product is mass-produced. This is followed by a number of commercial activities through which the product is actually sold in the market. Automation also reduces the over all product life cycle i.e., the time required to complete (i) Product conception and design (ii) Process planning and installation (iii) Various stages of the product life cycle are shown as in Figure 1.5.

Version 2 EE IIT, Kharagpur 8

Research and Development

Product Design

Market Feedbacks

Process Planning, Installation

Production, Quality Control

Fig. 1.5 A Typical Industrial Product Life Cycle

Economy of Scale and Economy of Scope
In the context of Industrial Manufacturing Automation, Economy of Scale is defined as follows.

Economy of Scale
Definition: Reduction in cost per unit resulting from increased production, realized through operational efficiencies. Economies of scale can be accomplished because as production increases, the cost of producing each additional unit falls. Obviously, Automation facilitates economy of scale, since, as explained above, it enables efficient large-scale production. In the modern industrial scenario however, another kind of economy, called the economy of scope assumes significance.

Economy of Scope
Definition : The situation that arises when the cost of being able manufacture multiple products simultaneously proves more efficient than that of being able manufacture single product at a time. Economy of scope arises in several sectors of manufacturing, but perhaps the most predominantly in electronic product manufacturing where complete product life cycle, from conception to market, are executed in a matter of months, if not weeks. Therefore, to shrink the time to market drastically use of automated tools is mandated in all phases of the product life cycle. Additionally, since a wide variety of products need to be manufactured within the life period of a factory, rapid programmability and reconfigurability of machines and processes becomes a key requirement for commercial success. Such an automated production system also enables the industry to exploit a much larger market and also protects itself against fluctuations in demand for a given class of products. Indeed it is being driven by the economy of scope, and Version 2 EE IIT, Kharagpur 9

enabled by Industrial Automation Technology that Flexible Manufacturing (i.e. producing various products with the same machine) has been conceived to increase the scope of manufacturing. Next let us see the various major kinds of production systems, or factories, exist. This would be followed by a discussion on the various types of automation systems that are appropriate for each of these categories.

Point to Ponder: 8
A. Can you give an example of an industry where economy of scope is more significant than the economy of scale? B. Can you give an example of an industry where economy of scale is more significant than the economy of scope? C. Can you give an example of an industry where both economy of scope, and economy of scale are significant?

Types of production systems
Major industrial processes can be categorized as follows based on their scale and scope of production. Continuous flow process: Manufactured product is in continuous quantities i.e., the product is not a discrete object. Moreover, for such processes, the volume of production is generally very high, while the product variation is relatively low. Typical examples of such processes include Oil Refineries, Iron and Steel Plants, Cement and Chemical Plants. Mass Manufacturing of Discrete Products: Products are discrete objects and manufactured in large volumes. Product variation is very limited. Typical examples are Appliances, Automobiles etc. Batch Production: In a batch production process the product is either discrete or continuous. However, the variation in product types is larger than in continuous-flow processes. The same set of equipment is used to manufacture all the product types. However for each batch of a given product type a distinct set of operating parameters must be established. This set is often referred to as the “recipe” for the batch. Typical examples here would be Pharmaceuticals, Casting Foundries, Plastic moulding, Printing etc. Job shop Production: Typically designed for manufacturing small quantities of discrete products, which are custom built, generally according to drawings supplied by customers. Any variation in the product can be made. Examples include Machine Shops, Prototyping facilities etc. The above types of production systems are shown in Figure 1.6 categorized according to volumes of production and variability in product types. In general, if the quantity of product is more there is little variation in the product and more varieties of product is manufactured if the quantity of product is lesser. Version 2 EE IIT, Kharagpur 10

Oil Refinery Iron & Steel Chemical Appliances Bicycles Foundry Food Processing

Continuous Flow Process Mass Manufacturing Of Discrete Products

Quantity

Batch Production

Machine Tools Prototypes

Job shop Production

Variety

Fig. 1.6 Types of Production Systems

Types of Automation Systems
Automation systems can be categorized based on the flexibility and level of integration in manufacturing process operations. Various automation systems can be classified as follows Fixed Automation: It is used in high volume production with dedicated equipment, which has a fixed set of operation and designed to be efficient for this set. Continuous flow and Discrete Mass Production systems use this automation. e.g. Distillation Process, Conveyors, Paint Shops, Transfer lines etc. A process using mechanized machinery to perform fixed and repetitive operations in order to produce a high volume of similar parts. Programmable Automation: It is used for a changeable sequence of operation and configuration of the machines using electronic controls. However, non-trivial programming effort may be needed to reprogram the machine or sequence of operations. Investment on programmable equipment is less, as production process is not changed frequently. It is typically used in Batch process where job variety is low and product volume is medium to high, and sometimes in mass production also. e.g. in Steel Rolling Mills, Paper Mills etc. Flexible Automation: It is used in Flexible Manufacturing Systems (FMS) which is invariably computer controlled. Human operators give high-level commands in the form of codes entered into computer identifying product and its location in the sequence and the lower level changes are done automatically. Each production machine receives settings/instructions from computer. These automatically loads/unloads required tools and carries out their processing instructions. After processing, products are automatically transferred to next machine. It is typically used in job shops and batch processes where Version 2 EE IIT, Kharagpur 11

product varieties are high and job volumes are medium to low. Such systems typically use Multi purpose CNC machines, Automated Guided Vehicles (AGV) etc. Integrated Automation: It denotes complete automation of a manufacturing plant, with all processes functioning under computer control and under coordination through digital information processing. It includes technologies such as computer-aided design and manufacturing, computer-aided process planning, computer numerical control machine tools, flexible machining systems, automated storage and retrieval systems, automated material handling systems such as robots and automated cranes and conveyors, computerized scheduling and production control. It may also integrate a business system through a common database. In other words, it symbolizes full integration of process and management operations using information and communication technologies. Typical examples of such technologies are seen in Advanced Process Automation Systems and Computer Integrated Manufacturing (CIM) As can be seen from above, from Fixed Automation to CIM the scope and complexity of automation systems are increasing. Degree of automation necessary for an individual manufacturing facility depends on manufacturing and assembly specifications, labor conditions and competitive pressure, labor cost and work requirements. One must remember that the investment on automation must be justified by the consequent increase in profitability. To exemplify, the appropriate contexts for Fixed and Flexible Automation are compared and contrasted. Fixed automation is appropriate in the following circumstances. A. Low variability in product type as also in size, shape, part count and material B. Predictable and stable demand for 2- to 5-year time period, so that manufacturing capacity requirement is also stable C. High production volume desired per unit time D. Significant cost pressures due to competitive market conditions. So automation systems should be tuned to perform optimally for the particular product. Flexible automation, on the other hand is used in the following situations. A. Significant variability in product type. Product mix requires a combination of different parts and products to be manufactured from the same production system B. Product life cycles are short. Frequent upgradation and design modifications alter production requirements C. Production volumes are moderate, and demand is not as predictable

Point to Ponder: 9
A. During a technical visit to an industry how can you identify the type of automation prevailing there from among the above types? B. For what kind of a factory would you recommend computer integrated manufacturing and why? C. What kind of automation would you recommend for manufacturing Version 2 EE IIT, Kharagpur 12

a. b. c. d. e. f. g.

Light bulbs Garments Textile Cement Printing Pharmaceuticals Toys

Lesson Summary
In this lesson we have dealt with the following topics: A. Definition of Automation and its relations with fields of Automatic Control and Information Technology: It is seen that both control and IT are used in automation systems to realize one or more of its functionalities. Also, while Control Technology is used for operation of the individual machines and equipment, IT is used for coordination, management and optimized operation of overall plants. B. The role played by Automation in realizing the basic goal of profitability of a manufacturing industry: It is seen that Automation can increase profitability in multiple ways by reducing labour, material and energy requirements, by improving quality as well as productivity. It is also seen that Automation is not only essential to achieve Economy of Scale, but also for Economy of Scope. C. Types of Factories and Automation Systems that are appropriate for them: Factories have been classified into four major categories based on the product volumes and product variety. Similarly Automation Systems are also categorized into four types and their appropriateness for the various categories of factories explained.

Exercises
A. Describe the role of Industrial Automation in ensuring overall profitability of a industrial production system. Be specific and answer point wise. Give examples as appropriate. B. State the main objectives of a modern industry (at least five) and explain the role of automation in helping achieve these. C. Explain with examples the terms “economy of scale” and “economy of scope”. How does industrial automation help in achieving these? Cite examples. D. Differentiate between a job shop and a flow shop with example what are their ‘process plant’ analogues? Give examples. E. Run any internet search engine and type “History of Automation” to prepare a term paper on the subject. F. There are some aspects of automation that have not been treated in the lesson. Consult references and prepare term papers on the impact of automation on a. Environmental Appropriateness for Industries b. Industrial Standardisation Certification such as ISO 9001 c. Industrial Safety Version 2 EE IIT, Kharagpur 13

G. Locate the major texts on Manufacturing Automation H. From the internet find alternate definitions of the terms : Industry, Automation and Control

Version 2 EE IIT, Kharagpur 14

Answers, Remarks and Hints to Points to Ponder Point to Ponder: 1
A. Why does an automated system achieve superior performance compared to a manual one? Ans: Because such systems can have more precision, more energy and more speed of operation than possible manually. Moreover using computing techniques, much more sophisticated and efficient operational solutions can be derived and applied in real-time. B. Can you give an example where this happens? Ans: This is the rule. Only few exceptions exist. How many of the millions of industrial products could be made manually?

Point to Ponder: 2
A. Can you explain the above definition in the context of a common control system, such as temperature control in an oven? Ans: Consider a temperature-controlled oven as found in many kitchens. A careful examination of the dials would show that one could control the temperature in the oven. This is a closed loop control operation. One can also control the time for which the oven is kept on. Note that in both cases the input signal to the process is the applied voltage to the heater coils. This input signal is varied as required to hold the temperature, by the controller. B. Is the definition applicable to open-loop as well as closed loop control? Ans: Yes

Point to Ponder: 3
C. Can you give an example of an automated system, which contains a control system as a part of it? Ans: Many examples can be given. One of these is the following: In an industrial CNC machine, the motion control of the spindle, the tool holder and the job table are controlled by a position and speed control system, which, in fact, uses a separate processor. Another processor is used to manage the other automation aspects. Another example is that of A pick and place automated robot is used in many industrial assembly shops. The robot motion can be programmed using a high level interface. The motion of the robot is controlled using position control systems driving the various joints in the robotic manipulator.

Version 2 EE IIT, Kharagpur 15

D. What are the other parts of the system? Ans: The other functional parts of the CNC System include: The operator interface, the discrete PLC controls of indicators, lubricant flow control, tool changing mechanisms.

Point to Ponder: 4
Try to find an example automated system which uses at least one of the areas of Industrial IT mentioned in Fig. 1.2. (Hint: Try using the internet) Ans: Distributed Control Systems (DCS) used in many large Continuous-Flow processes such as Petroleum Refining and Integrated Steel Plants use almost all components of Industrial IT

Point to Ponder: 5
A. Can you give an example of an automated system, some of whose parts makes a significant application of Industrial IT? Ans: Distributed Control Systems (DCS) used in many large Continuous-Flow processes such as Petroleum Refining and Integrated Steel Plants use almost all components of Industrial IT B. Can you give an example of an automated system, none of whose parts makes a significant application of Industrial IT? Ans: An automated conveyor system used in many large Discrete Manufacturing Plants such as bottled Beverage Plants use no components of Industrial IT.

Point to Ponder: 6
A. Can you give an example of an automated system, which is reactive in the sense mentioned above? Ans: Any feedback controller, such as an industrial PID controller is reactive since it interacts with sensors and actuators. B. Can you give an example of an automated system, which is real-time in the sense mentioned above Ans: Any feedback controller, such as an industrial PID controller is real-time, since it has to compute its output within one sampling time. C. Can you give an example of an automated system, which is mission-critical in the sense mentioned above

Version 2 EE IIT, Kharagpur 16

Ans: An automation system for a Nuclear Power Plant is mission critical since a failure is unacceptable for such a system.

Point to Ponder: 7
A. With reference to Eq. (1), explain how the following automation systems improve industrial profitability. d. Automated Welding Robots for Cars e. Automated PCB Assembly Machines f. Distributed Control Systems for Petroleum Refineries Ans: Some of the factors that lead to profitability in each case, are mentioned. a. Automated Welding Robots for Cars Increased production rate, Uniform and accurate welding, Operator safety. b. Automated PCB Assembly Machines Increased production rate, Uniform and accurate placement and soldering c. Distributed Control Systems for Petroleum Refineries Energy efficiency, Improved product quality

Point to Ponder: 8
A. You give an example of an industry where economy of scope is more significant than the economy of scale? Ans: One such example would a job shop which manufactures custom machine parts by machining according to customer drawings. Another example would be a factory to manufacture Personal Computer components B. Can you give an example of an industry where economy of scale is more significant than the economy of scope? Ans: One such example would be a Power plant. Another one would be a Steel Plant.

Point to Ponder: 9
A. During a technical visit to an industry how can you identify the type of automation prevailing there from among the above types? Ans: Check for the following. ♦ Whether automatic control exists for majority the equipment ♦ Whether supervisory control is manual, partially automated or largely automated ♦ Whether operator interfaces are computer integrated or not. ♦ Whether communication with individual control units can be done from supervisory interfaces through computers or not

Version 2 EE IIT, Kharagpur 17

♦ Whether any information network exists, to which automation system and controllers are connected ♦ Product variety, product volumes, batch sizes etc. ♦ Whether the material handling systems are automated and if so to what extent. The type of automation system can be determined based on these information, as discussed in the lesson. B. For what kind of a factory would you recommend computer integrated manufacturing and why? Ans: For large systems producing sophisticated and expensive products in large volumes having many subunits to be integrated in complex ways. C. What kind of automation would you recommend for manufacturing a. Light bulbs Ans: Fixed b. Garments Ans: Flexible c. Textile Ans: Programmable d. Cement Ans: Programmable e. Printing Ans: Flexible f. Pharmaceuticals Ans: Flexible g. Toys Ans: Flexible

Version 2 EE IIT, Kharagpur 18

Module 1
Introduction
Version 2 EE IIT, Kharagpur 1

Lesson 2
Architecture of Industrial Automation Systems
Version 2 EE IIT, Kharagpur 2

Lesson Objectives
• • • • To describe the various elements of an Industrial Automation Systems and how they are organized hierarchically in levels. To explain how these levels relate to each other in terms of their functions. To describe the nature of technologies involved in realizing these functional levels To describe the nature of information processing in these levels and the information flow among them

The Functional Elements of Industrial Automation
An Industrial Automation System consists of numerous elements that perform a variety of functions related to Instrumentation, Control, Supervision and Operations Management related to the industrial process. These elements may also communicate with one another to exchange information necessary for overall coordination and optimized operation of the plant/factory/process. Below, we classify the major functional elements typically found in IA systems and also describe the nature of technologies that are employed to realize the functions.

Sensing and Actuation Elements
These elements interface directly and physically to the process equipment and machines. The sensing elements translate the physical process signals such as temperature, pressure or displacement to convenient electrical or pneumatic forms of information, so that these signals can be used for analysis, decisions and finally, computation of control inputs. These computed control inputs, which again are in convenient electrical or pneumatic forms of information, need to be converted to physical process inputs such as, heat, force or flow-rate, before they can be applied to effect the desired changes in the process outputs. Such physical control inputs are provided by the actuation elements.

Industrial Sensors and Instrument Systems
Scientific and engineering sensors and instrument systems of a spectacular variety of size, weight, cost, complexity and technology are used in the modern industry. However, a close look would reveal that all of them are composed of a set of typical functional elements connected in a specified way to provide signal in a form necessary. The various tasks involved in the automation systems. Fig 2.1 below shows the configuration of a typical sensor system.
Sensing element Signal Conditioning element Signal Processing element Target Signal Handling element

Medium

Fig. 2.1 Functional configuration of a typical sensor system In Fig. 2.1 a sensor system is shown decomposed into three of its major functional components, along with the medium in which the measurement takes place. These are described below. Version 2 EE IIT, Kharagpur 3

A. The physical medium refers to the object where a physical phenomenon is taking place and we are interested in the measurement of some physical variable associated with the phenomenon. Thus, for example, the physical medium may stand for the hotga in a furnace in the case of temperature measurement or the fluid in a pipe section in the case of measurement of liquid flow rate. B. The sensing element is affected by the phenomenon in the physical medium either through direct or physical contact or through indirect interaction of the phenomenon in the medium with some component of the sensing element. Again, considering the case of temperature measurement, one may use a thermocouple probe as the sensing element that often comes in physical contact with the hot object such as the flue gas out of a boilerfurnace or an optical pyrometer which compares the brightness of a hot body in the furnace with that of a lamp from a distance through some window and does not come in direct contact with the furnace. In the more common case where the sensing element comes in contact with the medium, often some physical or chemical property of the sensor changes in response to the measurement variable. This change then becomes a measure of the physical variable of interest. A typical example is the change in resistivity due to heat in a resistance thermometer wire. Alternatively, in some other sensors a signal is directly generated in the sensing element, as is the case of a thermocouple that generates a voltage in response to a difference in temperature between its two ends. C. The signal-conditioning element serves the function of altering the nature of the signal generated by the sensing element. Since the method of converting the nature of the signal generated in the sensor to another suitable signal form (usually electrical) depends essentially on the sensor, individual signal conditioning modules are characteristic of a group of sensing elements. As an example consider a resistance Temperature Detector (RTD) whose output response is a change in its resistance due to change in temperature of its environment. This change in resistance can easily be converted to a voltage signal by incorporating the RTD in one arm of a Wheatstone's bridge. The bridge therefore serves as a signal-conditioning module. Signal conditioning modules are also used for special purpose functions relating to specific sensors but not related to variable conversion such as `ambient referencing' of thermocouples. These typically involve analog electronic circuits that finally produce electrical signals in the form of voltage or current in specific ranges. D. The signal processing element is used to process the signal generated by the first stage for a variety of purposes such as, filtering (to remove noise), diagnostics (to assess the health of the sensor), linearisation (to obtain an output which is linearly related with the physical measurand etc. Signal processing systems are therefore usually more general purpose in nature. E. The target signal-handling element may perform a variety of functions depending on the target application. It may therefore contain data/signal display modules, recording or/storage modules, or simply a feedback to a process control system. Examples include a temperature chart recorder, an instrumentation tape recorder, a digital display or an Analog to Digital Converter (ADC) followed by an interface to a process control computer. While the above description fits in most cases, it may be possible to discover some variations in some cases. The above separation into subsystems is not only from a functional point of view, Version 2 EE IIT, Kharagpur 4

more often than not, these subsystems are clearly distinguishable physically in a measurement system. Modern sensors often have the additional capability of digital communication using serial, parallel or network communication protocols. Such sensors are called “smart” and contain embedded digital electronic processing systems.

Point to Ponder: 1
A. Draw the functional block diagram of a typical sensor system B. Consider a strain-gage weigh bridge. Explore and identify the subsystems of the bridge and categorise these subsystems into the above mentioned classes of elements mentioned above.

Industrial Actuator Systems
Actuation systems convert the input signals computed by the control systems into forms that can be applied to the actual process and would produce the desired variations in the process physical variables. In the same way as in sensors but in a reverse sense, these systems convert the controller output, which is essentially information without the power, and in the form of electrical voltages (or at times pneumatic pressure) in two ways. Firstly it converts the form of the variable into the appropriate physical variable, such as torque, heat or flow. Secondly it amplifies the energy level of the signal manifold to be able to causes changes in the process variables. Thus, while both sensors and actuators cause variable conversions, actuators are high power devices while sensors are not. It turns out that in most cases, actuators are devices that first produce motion from electrical signal, which is then further converted to other forms. Based on the above requirement of energy and variable conversion most actuation systems are are structured as shown in Fig. 2.2.
Signal Processing element Power Amplifying Element Variable Conversion Element Energy Conversion Element

Process

Fig. 2.2 Functional configuration of a typical actuator system In Fig.2.2 an actuator system is shown decomposed into its major functional components, The salient points about the structure are described below. A. The electronic signal-processing element accepts the command from the control system in electrical form. The command is processed in various ways. For example it may be filtered to avoid applying input signals of certain frequencies that may cause resonance. Many actuators are themselves closed feedback controlled units for precision of the actuation operation. Therefore the electronic signal-processing unit often contains the control system for the actuator itself.

Version 2 EE IIT, Kharagpur 5

B. The electronic power amplification element sometimes contains linear power amplification stages called servo-amplifiers. In other cases, it may comprise power electronic drive circuits such as for motor driven actuators. C. The variable conversion element serves the function of altering the nature of the signal generated by the electronic power amplification element from electrical to non-electrical form, generally in the form of motion. Examples include electrohydraulic servo valve, stepper/servo motors, Current to Pneumatic Pressure converters etc. D. The non-electrical power conversion elements are used to amplify power further, if necessary, typically using hydraulic or pneumatic mechanisms. E. The non-electrical variable conversion elements may be used further to tranform the actuated variable in desired forms, often in several stages. Typical examples include motion-to-flow rate conversion in flow-valves, rotary to linear motion converters using mechanisms, flow-rate to heat conversion using steam or other hot fluids etc. F. Other Miscellaneous Elements such as Auxiliaries for Lubrication/Cooling/Filtering, Reservoirs, Prime Movers etc., sensors for feedback, components for display, remote operations, as well as safety mechanisms since the power handling level is significantly high.

Point to Ponder: 2
A. Draw the functional block diagram of an actuator system B. Consider an electro-hydraulic servo-valve actuator. Explore and identify the subsystems of the actuator and categorise these subsystems into the above-mentioned classes of elements mentioned above.

Industrial Control Systems
By industrial control systems, we denote the sensors systems, actuator systems as a controller. Controllers are essentially (predominantly electronic, at times pneumatic/hydraulic) elements that accept command signals from human operators or Supervisory Systems, as well as feedback from the process sensors and produce or compute signals that are fed to the actuators. Control Systems can be classified into two kinds.

Continuous Control
This is also often termed as Automatic Control, Process Control, Feedback Control etc. Here the controller objective is to provide such inputs to the plant such that the output y(t) follows the input r(t) as closely as possible, in value and over time. The structure of the common control loop with its constituent elements, namely the Controller, the Actuator, the Sensor and the Process itself is shown. In addition the signals that exist at various points of the

Version 2 EE IIT, Kharagpur 6

Disturbances di do uP

r (t)

+

e

Controller GC(s)

uA

Actuator GA(s)

Plant GP(s)

y(t) Output

Command/ reference/ Setpoint ds + +
Sensor GS(s)

Fig. 2.3 Typical control loop system are also marked. These include the command (alternatively termed the set point or the reference signal), the exogenous inputs (disturbances, noise). The difficulties in achieving the performance objective is mainly due to the unavoidable disturbances due to load variation and other external factors, as well as sensor noise, the complexity, possible instability, uncertainty and variability in the plant dynamics, as well as limitations in actuator capabilities. Most industrial control loop command signals are piecewise constant signals that indicate desirable levels of process variables, such as temperature, pressure, flow, level etc., which ensure the quality of the product in Continuous Processes. In some cases, such as in case of motion control for machining, the command signal may be continuously varying according to the dimensions of the product. Therefore, here deviation of the output from the command signal results in degradation of product quality. It is for this reason that the choice of the feedback signals, that of the controller algorithm (such as, P, PI pr PID), the choice of the control loop structure (normal feedback loop, cascade loop or feedforward) as well as choice of the controller gains is extremely important for industrial machines and processes. Typically the control configurations are well known for a given class of process, however, the choice of controller gains have to be made from time to time, since the plant operating characteristics changes with time. This is generally called controller tuning. A single physical device may act as the controller for one or more control loops (singleloop/multi-loop controller). Today, many loop controllers supplement typical control laws such as PID control by offering adaptive control and fuzzy logic algorithms to enhance controller response and operation. PID and startup self-tuning are among the most important features. Among other desired and commonly found characteristics are, ability to communicate upward with supervisory systems, as well as on peer-to-peer networks (such as Fieldbus or DeviceNet), support for manual control in the event of a failure in the automation. Software is an important factor in loop controllers. Set-up, monitoring and auto-tuning and alarm software for loop controllers is now a common feature. The controllers also accept direct interfacing of process sensors and signals. Choice of inputs includes various types of thermocouples, RTDs, voltage to 10 V dc, or current to 20 mA. While most sophisticated controllers today are electronic, Version 2 EE IIT, Kharagpur 7

pneumatic controllers are still being used. Pneumatic controllers are easy to use, easy to maintain, and virtually indestructible.

Point to Ponder: 3
A. Draw the block diagram of a typical industrial control system B. Consider a motor driven position control system, as commonly found in CNC Machine drives. Identify the main feedback sensors in the system. Identify the major sources of disturbance. How is such a drive different from that of an automated conveyor system?

Sequence / Logic Control
Many control applications do not involve analog process variables, that is, the ones which can assume a continuous range of values, but instead variables that are set valued, that is they only assume values belonging to a finite set. The simplest examples of such variables are binary variables, that can have either of two possible values, (such as 1 or 0, on or off, open or closed etc.). These control systems operate by turning on and off switches, motors, valves, and other devices in response to operating conditions and as a function of time. Such systems are referred to as sequence/logic control systems. For example, in the operation of transfer lines and automated assembly machines, sequence control is used to coordinate the various actions of the production system (e.g., transfer of parts, changing of the tool, feeding of the metal cutting tool, etc.). There are many industrial actuators which have set of command inputs. The control inputs to these devices only belong to a specific discrete set. For example in the control of a conveyor system, analog motor control is not applied. Simple on-off control is adequate. Therefore for this application, the motor-starter actuation system may be considered as discrete having three modes, namely, start, stop and run. Other examples of such actuators are solenoid valves, discussed in a subsequent lesson. Similarly, there are many industrial sensors (such as, Limit Switch / Pressure Switch/ Photo Switch etc.) which provide discrete outputs which may be interpreted as the presence/absence of an object in close proximity, passing of parts on a conveyor, or a given pressure value being higher or lower than a set value. These sensors thus indicate, not the value of a process variable, but the particular range of values to which the process variable belongs. A modern controller device used extensively for sequence control today in transfer lines, robotics, process control, and many other automated systems is the Programmable Logic Controller (PLC). In essence, a PLC is a special purpose industrial microprocessor based realtime computing system, which performs the following functions in the context of industrial operations

Point to Ponder: 4
A. State the major aspect in which sequence/logic control systems differ from analog control systems B. Describe an industrial system that employs discrete sensors and discrete actuators.

Version 2 EE IIT, Kharagpur 8

Supervisory Control
Supervisory control performs at a hierarchically higher level over the automatic controllers, which controls smaller subsystems. Supervisory control systems perform, typically the following functions: ♦ Set point computation: Set points for important process variables are computed depending on factors such as nature of the product, production volume, mode of processing. This function has a lot of impact on production volume, energy and quality and efficiency. ♦ Performance Monitoring / Diagnostics: Process variables are monitored to check for possible system component failure, control loop detuning, actuator saturation, process parameter change etc. The results are displayed and possibly archived for subsequent analysis. ♦ Start up / Shut down / Emergency Operations : Special discrete and continuous control modes are initiated to carry out the intended operation, either in response to operator commands or in response to diagnostic events such as detected failure modes. ♦ Control Reconfiguration / Tuning: Structural or Parametric redesign of control loops are carried out, either in response to operator commands or in response to diagnostic events such as detected failure modes. Control reconfigurations may also be necessary to accommodate variation of feedback or energy input e.g. gas fired to oil fired. ♦ Operator Interface: Graphical interfaces for supervisory operators are provided, for manual supervision and intervention. Naturally, these systems are dependent on specific application processes, in contrast with automatic control algorithms, which are usually generic (e.g. PID). Computationally these are a mixture of hard and soft real time algorithms. These are also often very expensive and based on proprietary knowledge of automating specific classes of industrial plants.

Point to Ponder: 5
A. State three major functions of a Supervisory Control System B. Consider the motor driven automatic position control system, as commonly found in CNC Machine drives. Explore and find out from where such systems get their set points during machining. Identify some of the other functionalities

Level 3: Production Control
Production control performs at a hierarchically higher level over the supervisory controllers. Typical functions they perform are: ♦ Process Scheduling: Depending on the sequence of operations to be carried on the existing batches of products, processing resource availability for optimal resource utilization. Version 2 EE IIT, Kharagpur 9

♦ Maintenance Management: Decision processes related to detection and deployment of maintenance operations. ♦ Inventory Management: Decision processes related to monitoring of inventory status of raw material, finished goods etc. and deployment of operations related to their management. ♦ Quality Management : Assessment, Documentation and Management of Quality Typically, the algorithms make use of Resource Optimisation Technology and are non-real-time although they may be using production data on-line.

Point to Ponder: 6
A. State three major functions of a Production Control System B. Explore and find out concrete activities for production control under at least two of the above major functions in any typical factory such as a Power Plans or a Steel Plant.

The Architecture of Elements: The Automation Pyramid
Industrial automation systems are very complex having large number of devices with confluence of technologies working in synchronization. In order to know the performance of the system we need to understand the various parts of the system. Industrial automation systems are organized hierarchially as shown in the following figure.

Level 4 Industrial IT Level 3 Enterprise Productio n Control Supervisory Control Automati c Control Sensors Actuator s Process / Fig. 2.4 Automation pyramid Version 2 EE IIT, Kharagpur 10

Level 2

Level 1 Industrial Auto Level 0

Various components in an industrial automation system can be explained using the automation pyramid as shown above. Here, various layers represent the wideness ( in the sense of no. of devices ), and fastness of components on the time-scale. Sensors and Acuators Layer: This layer is closest to the proceses and machines, used to translate signals sothat signals can be derived from processes for analysis and decisions and hence control signals can be applied to the processes. This forms the base layer of the pyramid also called ‘level 0’ layer. Automatic Control Layer: This layer consists of automatic control and monitoring systems, which drive the actuators using the process information given by sensors. This is called as ‘level 1’ layer. Supervisory Control Layer: This layer drives the automatic control system by setting target/goal to the controller. Supervisory Control looks after the equipment, which may consis of several control loops. This is called as ‘level 2’ layer. Production Control Layer: This solves the decision problems like production targets, resource allocation, task allocation to machines, maintenance management etc. This is called ‘level 3’ layer. Enterprise control layer: This deals less technical and more commercial activities like supply, demand, cash flow, product marketing etc. This is called as the ‘level 4’ layer. The spatial scale increases as the level is increased e.g. at lowest level a sensor works in a single loop, but there exists many sensors in an automation system which will be visible as the level is increased. The lowest level is faster in the time scale and the higher levels are slower. The aggregation of information over some time interval is taken at higher levels. All the above layers are connected by various types of communication systems. For example the sensors and actuators may be connected to the automatic controllers using a point-to-point digital communication, while the automatic controllers themselves may be connected with the supervisory and production control systems using computer networks. Some of these networks may be proprietary. Over the last decade, with emergence of embedded electronics and computing, standards for low level network standards (CANBus, Fieldbus etc.) for communication with low level devices, such as sensors and actuators are also emerging.

Point to Ponder: 7
A. Draw the Automation Pyramid and identify the Layers B. Give examples of the above major functional layers in any typical factor. . A concrete example of the Automation Functionality in a large manufacturing plant is presented in the appendix below. The appendix reveals the nature of functionality expected in modern automation systems, the elements that are used to realise them, and the figures of merit for such systems. The learner is encouraged to study it.

Version 2 EE IIT, Kharagpur 11

Appendix: An Example Industrial Specification for Automatic and Supervisory Level Automation Systems
This appendix contains the specification of a section of a Cold Rolling Mill complex, referred to here as PL-TCM which stands for Pickling Line and Tandem Control Mill. Such specification documents are prepared when automation systems for industrial plants are procured and installed. The document captures the visualisation of automation functionality of the customer. Here basic level refers to the automatic control supervisory control levels, while process control level refers to a level. Some of the terms and concepts described below have been discussed in subsequent lessons. Platforms: The above levels of controls shall be achieved through programmable controllers PLCs, micro-processor based systems as well as PCs / Work stations, as required. Each of the automation systems of the PL-TCM shall be subdivided in accordance with the functional requirements and shall cover the open loop and closed loop control functions of the different sections of the line and the mill. Modes of Operation: The systems shall basically have two modes of operation. In the semiautomatic mode the set point values shall be entered manually for different sections of the line through VDU and the processors shall transmit these values to the controls in proper time sequence. In fully automatic mode the process control system shall calculate all set point values through mathematical models and transfer the same to the subordinate systems over data link. The functions to be performed by the basic level automation shall cover but not be limited to the following. Functionality at Basic Level: The Basic Level shall cover control of all equipment, sequencing, interlocking micro-tracking of strip for specific functions, dedicated technological functions, storage of rolling schedules and look-up tables, fault and event logging etc. Some of these are mentioned below. ♦ All interlocking and sequencing control of the machinery such as for entry and exit handling of strips, shear control etc. Interlocking, sequencing, switching controls of the machines. This shall also cover automatic coil handling at the entry and exit sides, automatic sequencial operation of welding/rewelding machine and strip threading sequence control as well as for acid regeneration plant. ♦ Calculation of coil diameter and width at the entry pay-off reels. ♦ Position control of coil ears for centrally placing of coils on the mandrels. ♦ Generation of master speed references for the line depending on operator's input and line conditions and down loading to drive control systems. ♦ Speed synchronising control of the drives, as required. ♦ Strip tenstion, position and catenary control through control of related drives and machinery.

Version 2 EE IIT, Kharagpur 12

♦ Initiation of centre position control for Power Operated Rolls, steering/dancer rolls; Looper car position control. Automatic pre-setting control, measurement and control of tension and elongation for tension leveller. Auto edge position control at tension reels if required. ♦ Control of entry shear for auto-cutting of off-gauge strip. ♦ Control of pickling parameters for correct pickling with varying speed of strip in the pickling section. ♦ Side trimmer automatic setting contro. ♦ Interlockings, sequencing and control of scrap baller, if provided. ♦ Auto calibration for position control/precision positioning shall be provided as necessary. ♦ Manual/Auto slowdown/stoppage of strip at weld point at tension leveller, side trimmer, mill and exit shear. ♦ Control of technological functions for tandem mill such as : o Automatic gauge control along with interst and tension control. o Shape control o Roll force control ♦ Storage of tandem mill rolling schedules, for the entire product mix and all possible variations. Suitable look-up tables as operators guidance for line/equipment setting. ♦ Automatic roll changing along with automatic spindle positioning. ♦ Constant pass line control based on roll wear as well as after roll change. ♦ Automatic control of rotary shear before tension rells. ♦ Automatic sequence control of inspection reel. ♦ Provision of manual slow down/stoppage of strip as well as chearing for `run' for inspection of defects at tension leveller, side trimmer entry and exit of the Tandem Mill throuth push button stations. ♦ Micro-tracking of strip and flying gauge change (set point change) for continuous operation with varying strip sizes. ♦ Setting up the mill either from the stored rollings schedule with facility for modification by the operator of down-loading from process control level system. ♦ Automatic control of in-line coil weighing, marking and circumferential banding after delivery tension reels. Supervisory Functions at Basic Level: Centralised supervisory and monitoring control system shall be provided under basic level automation with dedicated processors and MMI. All necessary signals shall be acquired through drive control system as well as directly from the sensors/instruments as, required. The system shall be capable of carrying out the following fuctions.

Version 2 EE IIT, Kharagpur 13

♦ Centralised switching and start up of various line drives and auxiliary systems through mimic displays. ♦ Status of plant drives and electrical equipment for displaying maintenance information. ♦ Monitoring and display of measured values for tandem mill main drives and other large capacity drives such as winding temperature, for alarm and trip conditions. ♦ Centralised switching and status indication of 33 kV and 6.6 kV switchboards. ♦ Display of single line diagram of 33 kV and 6.6 kV switchboards, main drives, in-line auxiliary drives etc. ♦ Acquisition of fault signals from various sections of the plant with facility for display and print-out of the fault messages in clear text.

Comprehensive diagnostic functions
Functionality at Process Control Level: The Process Control Level shall be responsible for computation and control for optimization of operation. Functions like set point generation using mathematical models, learning control, material tracking within the process line/unit including primary data input, real time control of process functions through basic level automation, generation of reports etc. shall be implemented through this level of automation. Some of the specific functions to be performed by the process control level automation are the following. ♦ Coil strip tracking inside the process line/unit by sensing punched holes at weld seams. ♦ Primary Data Input (PDI) of coils at entry to PL-TCM with provision for down loading of data from production control level. ♦ Generation of all operating set points for the mill using PDI data, mill model, roll force model, power model, strip thickness control model, shape/profile control model with thermal strip flatness control as well as for other sections of the line. ♦ Learning (Adaptive) control using actual data and the mathematical model for set-up calculations. ♦ Storage of position setting values of levellers, side trimmer. Input of strip flaw data manually through inspection panel at the inline inspection facility after side trimmer. ♦ Processing of actual data on rolling operation, generation of reports logs and sending data to production control level. Information System Functions: The information system shall generally comply with the following features. ♦ Data of importance shall be available with the concerned personnel in the form of logs and reports. ♦ Output of logs and reports at preset times or on occurance of certain events. ♦ It shall be possible to change the data items and log formats without undue interference to the system. Version 2 EE IIT, Kharagpur 14

♦ Logged information shall be stored for adequate time period ensuring the availability of historical data record. ♦ Data captured by the system shall be checked for integrity with respect to their validity and plausibility with annunciation. Man Machine Interface: The visualisation system for both the automation levels shall be through man-machine interface (MMI) for the control and operation of the complete line. The system shall display the following screens, with facilities for hard copy print out. ♦ Process mimics for the complete line using various screens with status information of all important in-line drives as well as the references and actual values of important parameters. ♦ Dynamic information’s in form of bar graph for indication of reference and actual values of important parameters. ♦ Screens providing trends of the important process variables. ♦ Acquisition of actual parameters (averaging/maximum/minimum) for the complete line, on coil to coil basis through weld seam tracking or TCM exit shear cut for the generation of logs on process/parameters and production. Standards: The programmable controllers and other micro processor based equipment offered shall generally be designed/structured, manufactured and tested in accordance with the guidelines laid down in IEC-1131 (Part 2) apart from the industry standards being adopted by the respective manufactures. Hardware: The hardware of each basic controller/equipment of a system will generally comprise main processing unit, memory units, stabilised power supply unit, necessary communication interface modules, auxiliary storage where required. I/O modules in the main equipment, remote I/O stations where required and the programming and debugging tool (PADT). The hardware and software structure shall be modular to meet wide range of technological requirements. I/Os shall be freely configurable depending on the requirement. The programming units shall preferable be lap-top type. Networking: The networking would conform to the following specifications. ♦ In each of the two automation levels, all the controllers of a system shall be connected as a node over suitable data bus forming a LAN system using standardised hardware and software. ♦ The LAN system shall be in line with ISO-Open system Interconnect. ♦ All drive level automation equipment shall be suitably linked with the basic level for effective data/signal exchange between the two levels. However, all the emergency and safety signals shall be directly hardwired to the respective controllers. ♦ Similarly, the LAN systems for the basic level and process control level shall be suitably linked through suitable bridge/interface for effective data/signal exchange. Provision shall also be made for interfacing suitably the process control level with the production level automation system specified in item . Version 2 EE IIT, Kharagpur 15

♦ The data highways shall be designed to be optimally loaded and the same shall be clearly indicated in the offer. ♦ The remote I/Os, the microprocessor based measuring instruments and the microprocessor based special machines like coil weighing, marking and circumferantial banding machines shall be connected over serial links with the respective controllers. ♦ The personal computers and work stations shall be connected as a LAN system of the corresponding level. Data and Visualisation: The following specifications would apply in respect of data security, validity and its proper visualisation. ♦ All the operator interfaces comprising colour VDU and keyboard as MMI for interacting with the respective system and located at strategic locations, shall be connected to the corresponding LAN system. ♦ Keylock/password shall be provided to prevent unauthorised entry. ♦ Entry validity and plansibility check shall also be incorporated. ♦ An Engineer's console comprising of necessary processor, color VDU, keyboard/mouse and a printer unit shall be provided for the automation systems. The console shall have necessary hardware and software of communicating with the LAN and shall have access to the complete system. Basic functions of this console shall be off-line data base configuration, programme development, documentation etc. Application Software: The application software shall be through functional block type software modules as well as high level language based software modules. The software shall be user friendly and provided with help functions etc. Only one type of programming language shall be used for the complete system. However, ladder type programming language may be used for simple logical functions. Only industrially debugged and tested software shall be provided.

Basis of System Selection
Future Expandibility: The selection of equipment, standard software and networking shall be such as to offer optimum flexibility for future expansion without affecting the system reliability. Fault Tolerance: The system shall be designed to operate in automatic or semi automatic mode of operation under failure conditions. Spare Capacity: The system shall have sufficient capacity to perform all functions as required. A minimum of 30 per cent of the total memory shall be kept unallocated for future use. Loading: The data highway shall be designed to be optimally loaded and the same shall be clearly indicated in the offer. Software Structure and Quality Programs: shall be in high level language that is effective and economical for the proposed system in respect of Modularisation, rate of coding, store usage and running time. The software structure of the system shall besuitably distributed/centralised for supervision and control of the related process areas following the state of the art architecture. Version 2 EE IIT, Kharagpur 16

Integration: The communication software shall be such that the systems shall be able to communicate independently among themselves as well as with the lower level Basic Control/Process control automation system, as required. Provision shall be made for interfacing the production control system with the higher level Business Computer system to be provided for the entire steel plant in future. Programmability: The information system shall generally be designed such that it shall be possible to change the data items and log formats without undue interference to the system. Data Integrity and Protection: Logged information shall be stored for adequate time period ensuring the availability of historical data record. Data captured by the system shall be checked for integrity with respect to their validity and plausibility with annunciation. Storing of essential data to be protected against corruption when the system loses power supply or during failure.

Point to Ponder: 8
A. State three major functions of Supervisory Control mentioned in the lesson that have also been mentioned in the Automation System for the PL-TCM. B. State three major figures of merit for an Automation System mentioned in the appendix PL-TCM.

Version 2 EE IIT, Kharagpur 17

Answers, Remarks and Hints to Points to Ponder Point to Ponder: 1
Draw the functional block diagram of a typical sensor system Ans: The diagram is given in Fig. 2.1.in the lesson. Consider a strain-gage weigh bridge. Explore and identify the subsystems of the bridge and categorise these subsystems into the above mentioned classes of elements mentioned above Ans: A strain gage weighbridge contains the weighing platform and pillar, which senses the weight and produces a proportional strain (sensing element 1). This strain is sensed by a strain-gage which produces a proportional change in resistance (sensing element 2). The gage is incorporated into a Wheatstone’s bridge circuit (Signal conditioning) which generates a proportional unbalanced voltage.

Point to Ponder: 2
A. Draw the functional block diagram of an actuator system Ans: The diagram is given in Fig. 2.1.in the lesson. B. Consider an electro-hydraulic servo-valve actuator. Explore and identify the subsystems of the actuator and categorise these subsystems into the above-mentioned classes of elements mentioned above. Ans: An electrohydraulic servo valve is driven by current through a solenoid, which moves the spool of the valve, by applying a voltage across it. The voltage is derived by an electronic controller (electronic signal processing element), which gives a voltage input that is amplified by a servo amplifier (electronic power amplification element). The force due to the current produces motion of the spool (variable conversion element), which is converted to pressure (non electrical power conversion element) within the servo valve and applied to the final control element. Miscellaneous elements, such hydraulic system auxiliaries, indicators etc. are also present

Point to Ponder: 3
A. Draw the block diagram of a typical industrial control system Ans: The diagram is given in Fig. 2.3 in the lesson. B. Consider a motor driven position control system, as commonly found in CNC Machine drives. Identify the main feedback sensors in the system. Identify the major sources of disturbance. How is such a drive different from that of an automated conveyor system? Ans: Many examples can be given. One of these is the following:

Version 2 EE IIT, Kharagpur 18

In an industrial CNC machine, the motion control of the spindle, the tool holder and the job table are controlled by a position and speed control system, which, in fact, uses a position sensor such as shaft angle encoder or resolver, speed sensors such analog or digital tachometers and current sensors such Hall-effect sensors. The major sources of disturbances are changes in load torque arising in the machine due to material inhomogenity, tool wear etc. While both drives use motors for creating displacements, conveyor drives have very little demand on position and speed accuracy requirements. On the contrary there are very stringent requirement on these in the case of the CNC Machine.

Point to Ponder: 4
State the major aspect in which sequence/logic control systems differ from analog control systems Ans: The two major aspects in which they differ is in the nature of sensor inputs and the actuator outputs. These are discrete elements in the case of logic Control (on-off, low-high-medium etc.) and continuous valued in case of analog control. Similarly for actuator output (motor start/stop). The controller outputs are generally functions of inputs and feedback. Describe an industrial system that employs discrete sensors and discrete actuators. Ans: There are many such systems. For example a diecasting process is shown below. This process example is dealt with further in other lessons.

Industrial Example
The die stamping process is shown in figure below. This process consists of a metal stamping die fixed to the end of a piston. The piston is extended to stamp a work piece and retracted to allow the work piece to be removed. The process has 2 actuators: an up solenoid and a down solenoid, which respectively control the hydraulics for the extension and retraction of the stamping piston and die. The process also has 2 sensors: an upper limit switch that indicates when the piston is fully retracted and a lower limit switch that indicates when the piston is fully extended. Lastly, the process has a master switch which is used to start the process and to shut it down.

Version 2 EE IIT, Kharagpur 19

Piston

Upper limit switch Up Solenoid Down Solenoid Die

Lower limit switch

The control computer for the process has 3 inputs (2 from the limit sensors and 1 from the master switch) and controls 2 outputs (1 to each actuator solenoid).

Point to Ponder: 5
A. State three major functions of a Supervisory Control System Ans: Three major functions are: A. Set point generation B. Process Monitoring C. Operator Interface B. Consider the motor driven automatic position control system, as commonly found in CNC Machine drives. Explore and find out from where such systems get their set points during machining. Identify some of the other functionalities Ans: A separate processor is used to manage the above supervisory control aspects.

Point to Ponder: 6
A. State three major functions of a Production Control System Ans: Mentioned in text, they are 1. Process Scheduling 2. Maintenance Management, 3. Inventory Management B. Explore and find out concrete activities for production control under at least two of the above major functions in any typical factory such as a Power Plants or a Steel Plant. Version 2 EE IIT, Kharagpur 20

Ans: Power plants do not have product variety. However, heavy maintenance activity goes on round the year. There is also significant inventory management for the Coal Yard.

Point to Ponder: 7
A. Draw the Automation Pyramid and identify the Layers Ans: The diagram is given in Fig. 2.4.in the lesson. B. Give examples of the above major functional layers in any typical factory. Ans: The answer to this question is given in detail in the appendix for a section of a large rolling mill.

Point to Ponder: 8
A. State three major functions of Supervisory Control mentioned in the lesson that have also been mentioned in the Automation System for the PL-TCM. Ans: The answer to this question is given in detail in the appendix for a section of a large rolling mill. In brief three major functions are : ♦ Centralised switching and start up ♦ Monitoring and display of measured values ♦ Comprehensive diagnostic functions B. State three major figures of merit for an Automation System mentioned in the appendix PL-TCM. Ans: Integration, Programmability, Fault tolerance

Version 2 EE IIT, Kharagpur 21

Module 2
Measurement Systems
Version 2 EE IIT, Kharagpur 1

Lesson 3
Measurement Systems Specifications
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
At the end of this lesson, the student will be able to 1. 2. 3. Define the different terms used for characterizing the performance of an instrument/ measurement system. Compare the performances of two similar type of instruments, looking at the specifications Write down the performance specifications of a measurement system from its test data.

Introduction
One of the most frequent tasks that an Engineer involved in the design, commissioning, testing, purchasing, operation or maintenance related to industrial processes, is to interpret manufacturer’s specifications for their own purpose. It is therefore of paramount importance that one understands the basic form of an instrument specification and at least the generic elements in it that appear in almost all instrument specifications. Different blocks of a measurement system have been discussed in lesson-2. The combined performance of all the blocks is described in the specifications. Specifications of an instrument are provided by different manufacturers in different wrap and quoting different terms, which sometimes may cause confusion. Moreover, there are several application specific issues. Still, broadly speaking, these specifications can be classified into three categories: (i) static characteristics, (b) dynamic characteristics and (iii) random characteristics.

1.

Static Characteristics

Static characteristics refer to the characteristics of the system when the input is either held constant or varying very slowly. The items that can be classified under the heading static characteristics are mainly:

Range (or span)
It defines the maximum and minimum values of the inputs or the outputs for which the instrument is recommended to use. For example, for a temperature measuring instrument the input range may be 100-500 oC and the output range may be 4-20 mA.

Sensitivity
It can be defined as the ratio of the incremental output and the incremental input. While defining the sensitivity, we assume that the input-output characteristic of the instrument is approximately linear in that range. Thus if the sensitivity of a thermocouple is denoted as 10 μ V / 0C , it indicates the sensitivity in the linear range of the thermocouple voltage vs. temperature characteristics. Similarly sensitivity of a spring balance can be expressed as 25 mm/kg (say), indicating additional load of 1 kg will cause additional displacement of the spring by 25mm.

Version 2 EE IIT, Kharagpur 3

Again sensitivity of an instrument may also vary with temperature or other external factors. This is known as sensitivity drift. Suppose the sensitivity of the spring balance mentioned above is 25 mm/kg at 20 oC and 27 mm/kg at 30oC. Then the sensitivity drift/oC is 0.2 (mm/kg)/oC. In order to avoid such sensitivity drift, sophisticated instruments are either kept at controlled temperature, or suitable in-built temperature compensation schemes are provided inside the instrument.

Linearity
Linearity is actually a measure of nonlinearity of the instrument. When we talk about sensitivity, we assume that the input/output characteristic of the instrument to be approximately linear. But in practice, it is normally nonlinear, as shown in Fig.1. The linearity is defined as the maximum deviation from the linear characteristics as a percentage of the full scale output. Thus, ΔO (1) Linearity = Omax − Omin where, ΔO = max(ΔO1 , ΔO2 ) . Output OMAX Output OMAX

ΔΟ1

ΔΟ2 OMIN IMIN Fig. 1 Linearity IMAX OMIN Input IMIN

H IMAX

Input

Fig. 2 Hysteresis

Hysteresis
Hysteresis exists not only in magnetic circuits, but in instruments also. For example, the deflection of a diaphragm type pressure gage may be different for the same pressure, but one for increasing and other for decreasing, as shown in Fig.2. The hysteresis is expressed as the maximum hysteresis as a full scale reading, i.e., referring fig.2, H (2) Hysteresis = X 100. Omax − Omin

Resolution
In some instruments, the output increases in discrete steps, for continuous increase in the input, as shown in Fig.3. It may be because of the finite graduations in the meter scale; or the Version 2 EE IIT, Kharagpur 4

1 instrument has a digital display, as a result the output indication changes discretely. A 3 -digit 2 voltmeter, operating in 0-2V range, can have maximum reading of 1.999V, and it cannot measure any change in voltage below 0.001V. Resolution indicates the minimum change in input variable that is detectable. For example, an eight-bit A/D converter with +5V input can measure 5 or 19.6 mv. Referring to fig.3, resolution is also defined in terms the minimum voltage of 8 2 −1 of percentage as: ΔI Resolution = (3) X 100 I max − I min The quotient between the measuring range and resolution is often expressed as dynamic range and is defined as: measurement range Dynamic range = (4) resolution And is expressed in terms of dB. The dynamic range of an n-bit ADC, comes out to be approximately 6n dB.
Output

ΔΙ
Imin Imax Fig. 3 Resolution Input

Accuracy
Accuracy indicates the closeness of the measured value with the actual or true value, and is expressed in the form of the maximum error (= measured value – true value) as a percentage of full scale reading. Thus, if the accuracy of a temperature indicator, with a full scale range of 0500 oC is specified as ± 0.5%, it indicates that the measured value will always be within ± 2.5 oC of the true value, if measured through a standard instrument during the process of calibration. But if it indicates a reading of 250 oC, the error will also be ± 2.5 oC, i.e. ± 1% of the reading. Thus it is always better to choose a scale of measurement where the input is near full-scale value. But the true value is always difficult to get. We use standard calibrated instruments in the laboratory for measuring true value if the variable.

Precision
Precision indicates the repeatability or reproducibility of an instrument (but does not indicate accuracy). If an instrument is used to measure the same input, but at different instants, spread
Version 2 EE IIT, Kharagpur 5

over the whole day, successive measurements may vary randomly. The random fluctuations of readings, (mostly with a Gaussian distribution) is often due to random variations of several other factors which have not been taken into account, while measuring the variable. A precision instrument indicates that the successive reading would be very close, or in other words, the standard deviation σ e of the set of measurements would be very small. Quantitatively, the precision can be expressed as: measured range (5) Precision = σe The difference between precision and accuracy needs to be understood carefully. Precision means repetition of successive readings, but it does not guarantee accuracy; successive readings may be close to each other, but far from the true value. On the other hand, an accurate instrument has to be precise also, since successive readings must be close to the true value (that is unique).

2.

Dynamic Characteristics

Dynamic characteristics refer to the performance of the instrument when the input variable is changing rapidly with time. For example, human eye cannot detect any event whose duration is more than one-tenth of a second; thus the dynamic performance of human eye cannot be said to be very satisfactory. The dynamic performance of an instrument is normally expressed by a differential equation relating the input and output quantities. It is always convenient to express the input-output dynamic characteristics in form of a linear differential equation. So, often a nonlinear mathematical model is linearised and expressed in the form:
d n x0 d n −1 x 0 dx d m xi d m −1 xi dx + a n −1 + ⋅ ⋅ ⋅ ⋅ + a1 0 + a 0 x 0 = bm + bm −1 + ⋅ ⋅ ⋅ ⋅ +b1 i + b0 xi n n −1 m m −1 dt dt dt dt dt dt (6) where xi and x 0 are the input and the output variables respectively. The above expression can also be expressed in terms of a transfer function, as: x ( s ) bm s m + bm −1 s m −1 ⋅ ⋅ ⋅ +b1 s + b0 (7) = G ( s) = 0 xi ( s ) a n s n + bn −1 s n −1 ⋅ ⋅ ⋅ + a1 s + a 0 Normally m<n an n is called the order of the system. Commonly available sensor characteristics can usually be approximated as either zero-th order, first order or second order dynamics. Here are few such examples: an

Potentiometer
Displacement sensors using potentiometric principle (Fig.4) have no energy storing elements. The output voltage eo can be related with the input displacement xi by an algebraic equation: e ( s) E = = constant eo (t ) xt = Exi (t ); or , o (8) xi ( s ) x t where xt is the total length of the potentiometer and E is the excitation voltage.. So, it can be termed as a zeroth order system.

Version 2 EE IIT, Kharagpur 6

Thermocouple
A bare thermocouple (Fig.5) has a mass (m) of the junction. If it is immersed in a fluid at a temperature Tf , then its dynamic performance relating the output voltage eo and the input temperature Tf , can be expressed by the transfer function: eo ( s ) Kv (9) = T f ( s) 1 + sτ where, K v = steady state voltage sensitivity of the thermocouple in V/ oC. mC τ = time constant of the thermocouple = hA m = mass of the junction C = specific heat h = heat transfer co-efficient A = surface area of the hot junction. Hence, the bare thermocouple is a first order sensor. But if the bare thermocouple is put inside a metallic protective well (as it is normally done for industrial thermocouples) the order of the system increases due to the additional energy storing element (thermal mass of the well) and it becomes a second order system.

Seismic Sensor
Seismic sensors (Fig.6.) are commonly used for vibration or acceleration measurement of foundations. The transfer function between the input displacement xi and output displacement

xo can be expressed as:
xo ( s) Ms 2 (10) = xi ( s) Ms 2 + Bs + K where: M = mass of the seismic body B = damping constant K= spring constant From the above transfer function, it can be easily concluded that the seismic sensor is a second order system. eo
+ xo M E xt xi eo K Tf B xi Fig. 6 Seismic sensor Version 2 EE IIT, Kharagpur 7

Fig. 4 Potentiometer

Fig. 5 Thermocouple

Dynamic characteristics specifications are normally referred to the referred to the performance of the instrument with different test signals, e.g. impulse input, step input, ramp input and sinusoidal input. Few important specifications are:

1 Normalised response

Mp 2%

ts 0 time Fig. 7 Step response of a dynamic system

Step response performance
The normalized step response of a measurement system normally encountered is shown in Fig. 7. Two important parameters for classifying the dynamic response are:
Peak Overshoot (Mp): It is the maximum value minus the steady state value, normally expressed in terms of percentage. Settling Time (ts): It is the time taken to attain the response within ± 2% of the steady state value. Rise time (tr): It is the time required for the response to rise from 10% to 90% of its final value.

Frequency Response Performance
The frequency response performance refers to the performance of the system subject to sinusoidal input of varying frequency.

A sinωt

Dynamic System G(s)

B sin(ωt + φ)

Fig. 8 Frequency Response

Version 2 EE IIT, Kharagpur 8

Suppose G(s) is the transfer function of the dynamic measurement system, represented by the general relation (7). If the input is a sinusoidal quantity of amplitude A and frequency ω , then in the steady state, the output will also be of same frequency, but of different amplitude B, and there would be a phase difference between the input and output. It can be shown that the amplitude ratio and the phase difference can be obtained as:
B = G ( jω ) A

and φ = ∠G ( jω )

(11)

Normalized Amplitude ratio

1 0.707

Bandwidth

10

102

104

106 ω

Fig. 9 Amplitude vs. frequency characteristics of a piezoelectric accelerometer

The plots showing variations of amplitude ratios and phase angle with frequency are called the magnitude and phase plots of the frequency response. Typical amplitude vs. frequency characteristics of a piezoelectric accelerometer is shown in Fig.9.

Bandwidth and Natural Frequency
From fig. 9, it is apparent that the amplitude is fairly constant over a range of frequencies. This range is called the bandwidth of the measuring system (to be precise, it is the frequency range in which the normalized amplitude ratio does not fall below 0.707, or -3 dB limit). The instrument is suitable for use in this range. The lower and upper limits are called the lower and upper cut off frequencies. The frequency at which the amplitude ratio attains a peak is called the (damped) natural frequency of the system. For further details, the reader is requested to refer any standard book on control systems.

3.

Random Characteristics

If repeated readings of the same quantity of the measurand are taken by the same instrument, under same ambient conditions, they are bound to differ from each other. This is often due to some inherent sources of errors of the instrument that vary randomly and at any point of time it is very difficult to exactly say, what would be its value. For example, the characteristics of resistance and diode elements of an electronic circuit are random, due to two sources of noises: thermal noise and flicker noise. To characterize these behaviors, statistical terminologies are Version 2 EE IIT, Kharagpur 9

often used. Most common among them are Mean and Standard deviation. The mean of a set of readings is the most accurate estimation of the actual value, since, the positive and negative errors often cancelled out. On the other hand the standard deviation ( σ ) is a measure of the spread of the readings. If successive measurements of the same parameter under same ambient condition are taken and the mean and standard deviations are calculated, then assuming normal distribution of the randomness in measurements, we can say that 68% of the readings would fall within the range of mean ± σ . Naturally, smaller the value of σ , more would be the repeatability and higher would be the precision (refer equation (5)). This uncertainty limit is often extended to the 3 σ limit, that means, that with 99% confidence, we can say that the any reading taken, would give a value within the range of mean ± 3 σ . The interval of uncertainty is often called as the confidence interval. This part is discussed in detail in Lesson10.

Exercise
1. Followings are the excerpts from the specifications of a laser displacement sensor: (a) Measurement range: ± 10mm (b) Measurement point: 40mm (c) Resolution: 3 μ m (d) Linearity: 1% Full Scale (e) Response time: 0.15ms (f) Linear output: 4-20mA

Answer the following questions
i. ii. iii. iv. 2. 3. 4. Explain the meaning of each term. Suppose, the distance between the sensor and the object is 35mm. Then what would be output in mA? What is the error due to nonlinearity under the above condition? Find out the sensitivity of the sensor in mA/mm.

Find the resolution of a 10-bit ADC, if it is excited by a 10V source. The accuracy specified for a pressure gauge of range 0-10kPa is 2%. Find the maximum error in measurement in Pa if it gives a reading of 4.0 kPa. Justify the following statements: (a) A potentiometer is a zero-th order device. (b) A bare thermocouple is a first order device. (c) An accelerometer is a second order device.

Answers
1. (ii) 8mA, (iii) ± 0.1mm, (iv) 0.8 mA/mm. 2. 9.77mv. 3. ± 0.2kPa.

Version 2 EE IIT, Kharagpur 10

Module 2
Measurement Systems
Version 2 EE IIT, Kharagpur 1

Lesson 4
Temperature Measurement
Version 2 EE IIT, Kharagpur 2

Instructional Objective
The reader, after going through the lesson would be able to 1. Name different methods for temperature measurement 2. Distinguish between the principles of operation of RTD and thermistor 3. Explain the meaning of lead wire compensation of RTD 4. Differentiate characteristics of a PTC thermistor from a NTC thermistor 5. Select the proper thermocouple for a particular temperature range 6. Design simple cold junction compensation schemes for thermocouples.

1.

Introduction

The word temper was used in the seventeenth century to describe the quality of steel. It seems, after the invention of crude from of thermometer, the word temperature was coined to describe the degree of hotness or coolness of a material body. It was the beginning of seventeenth century when the thermometer – a temperature measuring instrument was first developed. Galileo Galilei is credited with the construction of first thermometer, although a Dutch scientist Drebbel also made similar instrument independently. The principle was simple. A bulb containing air with long vertical tube was inverted and dipped into a basin of water or coloured liquid. With the change in temperature of the bulb, the gas inside expanded or contracted, thus changing the level of the liquid column inside the vertical tube. A major drawback of the instrument was that it was sensitive not only to variation of temperature, but also to atmospheric pressure variation. Successive developments of thermometers came out throughout seventeenth and eighteenth century. The liquid thermometer was developed during this time. The importance of two reference fixed temperatures was felt while graduating the temperature scales. Boiling point of water and melting point of ice provided two easily available references. But some other references were also tried. Fahrenheit developed a thermometer where, it seems, temperature of ice and salt mixture was taken as 0o and temperature of human body as 96o. These two formed the reference points, with which, the temperature of melting ice came as 32o and that of boiling water as 212o. In Celsius scale, the melting point of ice was chosen as 0o and boiling point of water as 100o. The concept of Kelvin scale came afterwards, where the absolute temperature of gas was taken as 00 and freezing point of water as 273o. The purpose of early thermometers was to measure the variation of atmospheric or body temperatures. With the advancement of science and technology, now we require temperature measurement over a wide range and different atmospheric conditions, and that too with high accuracy and precision. To cater these varied requirements, temperature sensors based on different principles have been developed. They can be broadly classified in the following groups: 1. Liquid and gas thermometer 2. Bimetallic strip 3. Resistance thermometers (RTD and Thermistors) 4. Thermocouple Version 2 EE IIT, Kharagpur 3

5. Junction semiconductor sensor 6. Radiation pyrometer Within the limited scope of this course, we shall discuss few of the above mentioned temperature sensors, that are useful for measurement in industrial environment.

2.

Resistance Thermometers

It is well known that resistance of metallic conductors increases with temperature, while that of semiconductors generally decreases with temperature. Resistance thermometers employing metallic conductors for temperature measurement are called Resistance Temperature Detector (RTD), and those employing semiconductors are termed as Thermistors. RTDs are more rugged and have more or less linear characteristics over a wide temperature range. On the other hand Thermistors have high temperature sensitivity, but nonlinear characteristics.

Resistance Temperature Detector
The variation of resistance of metals with temperature is normally modeled in the form: R t = R 0 [1 + α( t − t 0 ) + β( t − t 0 ) 2 + ......] (1) where R t and R 0 are the resistance values at to C and t0oC respectively; α, β, etc. are constants that depends on the metal. For a small range of temperature, the expression can be approximated as: R t = R 0 [1 + α( t − t 0 )] (2) For Copper, α = 0.00427 / o C . Copper, Nickel and Platinum are mostly used as RTD materials. The range of temperature measurement is decided by the region, where the resistance-temperature characteristics are approximately linear. The resistance versus temperature characteristics of these materials is shown in fig.1, with to as 0oC. Platinum has a linear range of operation upto 650oC, while the useful range for Copper and Nickel are 120oC and 300oC respectively.

Version 2 EE IIT, Kharagpur

4

6 5 4 Rt/R0 3 2 1 Nickel Copper Platinum 0 200 400 600

Temperature (ºC) Fig. 1 Resistance-temperature characteristics of metals

Construction
For industrial use, bare metal wires cannot be used for temperature measurement. They must be protected from mechanical hazards such as material decomposition, tearing and other physical damages. The salient features of construction of an industrial RTD are as follows: • The resistance wire is often put in a stainless steel well for protection against mechanical hazards. This is also useful from the point of view of maintenance, since a defective sensor can be replaced by a good one while the plant is in operation. Heat conducting but electrical insulating materials like mica is placed in between the well and the resistance material. The resistance wire should be carefully wound over mica sheet so that no strain is developed due to length expansion of the wire.

• •

Fig. 2 shows the cut away view of an industrial RTD.

Version 2 EE IIT, Kharagpur

5

Stainless steel well

Resistance wire Mica Ceramic powder Fig. 2 Construction of an industrial RTD

Signal conditioning
The resistance variation of the RTD can be measured by a bridge, or directly by volt-ampere method. But the major constraint is the contribution of the lead wires in the overall resistance measured. Since the length of the lead wire may vary, this may give a false reading in the temperature to be measured. There must be some method for compensation so that the effect of lead wires is resistance measured is eliminated. This can be achieved by using either a three wire RTD, or a four wire RTD. Both the schemes of measurement are shown in fig. 3. In three wire method one additional dummy wire taken from the resistance element and connected in a bridge (fig. 3(a)) so that the two lead wires are connected to two adjacent arms of the bridge, thus canceling each other’s effect. In fig. 3(b) the four wire method of measurement is shown. It is similar to a four terminal resistance and two terminals are used for injecting current, while two others are for measuring voltage.

a b V c d a b RX c a, b, c: lead wires Fig. 3(b) Four wire RTD a, b, c: lead wires Fig. 3(a) Three wire RTD

Version 2 EE IIT, Kharagpur

6

Thermistor
Thermistors are semiconductor type resistance thermometers. They have very high sensitivity but highly nonlinear characteristics. This can be understood from the fact that for a typical 2000 Ω the resistance change at 25oC is 80Ω/oC, whereas for a 2000 Ω platinum RTD the change in resistance at 25oC is 7Ω/oC. Thermistors can be of two types: (a) Negative temperature coefficient (NTC) thermistors and (b) Positive temperature co-efficient (PTC) thermistors. Their resistance-temperature characteristics are shown in fig. 4(a) and 4(b) respectively. The NTC thermistors, whose characteristics are shown in fig. 4(a) is more common. Essentially, they are made from oxides of iron, manganese, magnesium etc. Their characteristics can be expressed as:

R T = R 0e
where,

1 1 β( − ) T T0

(3)

RT is the resistance at temperature T (K) R0 is the resistance at temperature T0 (K) T0 is the reference temperature, normally 25oC β is a constant, its value is decided by the characteristics of the material, the nominal value is taken as 4000. From (3), the resistance temperature co-efficient can be obtained as: 1 dR T β αT = =− 2 (4) R T dT T It is clear from the above expression that the negative sign of α T indicates the negative resistance-temperature characteristics of the NTC thermistor.

Resistance

Resistance

Temperature Fig. 4(a) Characteristics of a NTC thermistor

TR Temperature Fig. 4(b) Characteristics of a PTC thermistor

Useful range of themistors is normally -100 to +300oC. A single thermistor is not suitable for the whole range of measurement. Moreover, existing thermistors are not interchangeable. There is a marked spread in nominal resistance and the temperature coefficient between two thermistors of same type. So, if a defective thermistor is to be replaced by a new thermistor similar type, a fresh
Version 2 EE IIT, Kharagpur 7

calibration has to be carried out before use. Commercially available thermistors have nominal values of 1K, 2K, 10K, 20K, 100K etc. The nominal values indicate the resistance value at 25oC. Thermistors are available in different forms: bead type, rod type disc type etc. The small size of the sensing element makes it suitable for measurement of temperature at a point. The time constant is also very small due to the small thermal mass involved. The nonlinear negative temperature characteristics also give rise to error due to self-heating effect. When current is flowing through the thermistor, the heat generated due to the I 2 R -loss may increase the temperature of the resistance element, which may further decrease the resistance and increase the current further. This effect, if not tackled properly, may damage the thermistor permanently. Essentially, the current flowing should be restricted below the specified value to prevent this damage. Alternatively, the thermistor may be excited by a constant current source. The nonlinear characteristics of thermistors often creates problem for temperature measurement, and it is often desired to linearise the thermistor characteristics. This can be done by adding one fixed resistance parallel to the thermistor. The resistance temperature characteristics of the equivalent resistance would be more linear, but at the cost of sensitivity. The Positive Temperature Coefficient (PTC) thermistor have limited use and they are particularly used for protection of motor and transformer widings. As shown in fig. 4(b), they have low and relatively constant resistance below a threshold temperature TR, beyond which the resistance increases rapidly. The PTC thermistors are made from compound of barium, lead and strontium titanate.

3.

Thermocouple

Thomas Johan Seeback discovered in 1821 that thermal energy can produce electric current. When two conductors made from dissimilar metals are connected forming two common junctions and the two junctions are exposed to two different temperatures, a net thermal emf is produced, the actual value being dependent on the materials used and the temperature difference between hot and cold junctions. The thermoelectric emf generated, in fact is due to the combination of two effects: Peltier effect and Thomson effect. A typical thermocouple junction is shown in fig. 5. The emf generated can be approximately expressed by the relationship:

e 0 = C1 (T1 − T2 ) + C 2 (T12 − T22 ) μv

(5)

where T1 and T2 are hot and cold junction temperatures in K. C1 and C2 are constants depending upon the materials. For Copper/ Constantan thermocouple, C1=62.1 and C2=0.045.

Version 2 EE IIT, Kharagpur

8

A

T1 (hot junction)

T2 B e0 Fig. 5 A typical thermocouple B (cold junction)

Thermocouples are extensively used for measurement of temperature in industrial situations. The major reasons behind their popularity are: (i) they are rugged and readings are consistent, (ii) they can measure over a wide range of temperature, and (iii) their characteristics are almost linear with an accuracy of about ± 0.05%. However, the major shortcoming of thermocouples is low sensitivity compared to other temperature measuring devices (e.g. RTD, Thermistor).

Thermocouple Materials
Theoretically, any pair of dissimilar materials can be used as a thermocouple. But in practice, only few materials have found applications for temperature measurement. The choice of materials is influenced by several factors, namely, sensitivity, stability in calibration, inertness in the operating atmosphere and reproducibility (i.e. the thermocouple can be replaced by a similar one without any recalibration). Table-I shows the common types of thermocouples, their types, composition, range, sensitivity etc. The upper range of the thermocouple is normally dependent on the atmosphere whre it has been put. For example, the upper range of Chromel/ Alumel thermocouple can be increased in oxidizing atmosphere, while the upper range of Iron/ Constantan thermocouple can be increased in reducing atmosphere.
Table-1 Thermocouple materials and Characteristics Type Positive lead Negative lead Temperature Temperature range coeff.variation μv/oC Most linear range and sensitivity in the range 1100-1500oC 13.6-14.1 μv/oC

R

S

PlatinumPlatinum Rhodium (87% Pt, 13% Rh) PlatinumPlatinum Rhodium (90% Pt, 10% Rh) Chromel (90%Ni, 10% Cr) Alumel (Ni94Al2 Mn3Si)

0-1500oC

5.25-14.1

0-1500oC

5.4-12.2

1100-1500 oC 13.6-14.1 μv/oC 0-1000 oC 38-42.9 μv/oC

K

-200-1300oC

15.2-42.6

Version 2 EE IIT, Kharagpur

9

E

Chromel

Constantan (57%Cu, 43%Ni)

-200-1000oC

25.1-80.8

300-800 oC 77.9-80.8 μv/oC nonlinear 100-500 oC 54.4-55.9

T J

Copper Iron

Constantan

-200-350oC

15.8-61.8 21.8-64.6

Constantan -150-750oC

Laws of Thermocouple
The Peltier and Thompson effects explain the basic principles of thermoelectric emf generation. But they are not sufficient for providing a suitable measuring technique at actual measuring situations. For this purpose, we have three laws of thermoelectric circuits that provide us useful practical tips for measurement of temperature. These laws are known as law of homogeneous circuit, law of intermediate metals and law of intermediate temperatures. These laws can be explained using fig. 6. The first law can be explained using fig. 6(a). It says that the net thermo-emf generated is dependent on the materials and the temperatures of two junctions only, not on any intermediate temperature. According to the second law, if a third material is introduced at any point (thus forming two additional junctions) it will not have any effect, if these two additional junctions remain at the same temperatures (fig. 6(b)). This law makes it possible to insert a measuring device without altering the thermo-emf.

Version 2 EE IIT, Kharagpur 10

T3 T1 B

T4 A

Same net emf
T5 T2 T1 B (a) T8

T9 A

T10 T2

+
T6

e
T7

B

+
T11

T12

B

A T1 B

Same net emf
T2 T1 B (b)

A

T2

e
T2

T3 C T4

e

B

A T1

A T2 B T2

A T3 B T1

+ e1 B

+ e2 B

+ e3 B

T3 B

e3 = e1 + e2
(c)

Fig. 6 Laws of Thermocouple

The third law is related to the calibration of the thermocouple. It says, if a thermocouple produces emf e1, when its junctions are at T1 and T2, and e2 when its junctions are at T2 and T3; then it will generate emf e1+e2 when the junction temperatures are at T1 and T3 (fig. 6(c)). The third law is particularly important from the point of view of reference junction compensation. The calibration chart of a thermocouple is prepared taking the cold or reference junction temperature as 0oC. But in actual measuring situation, seldom the reference junction temperature is kept at that temperature, it is normally kept at ambient temperature. The third law helps us to compute the actual temperature using the calibration chart. This can be explained from the following example.

Example-1
The following table has been prepared from the calibration chart of iron-constantan thermocouple, with reference temperature at 0oC.

Version 2 EE IIT, Kharagpur 11

15 30 Temperature (oC) emf (mv) 0.778 1.56

40 2.11

.…….. ………

180 9.64

190

200

208 11.20

210 11.32

10.25 10.74

Suppose, the temperature of the hot junction is measured with a iron-constantan thermocouple, with the reference junction temperature at 30oC. If the voltage measured is 9.64mv, find the actual temperature of the hot junction.

Solution
Referring fig. 6(c), for this problem, T1 is the unknown temperature, T2= 30oC, T3= 0oC. The voltages are e1=9.64mv (measured) and e2= 1.56mv (from chart). Therefore, e3= e1+e2= 11.20mv. Hence, from the calibration chart, the actual hot junction temperature is T1 = 208oC.

Reference Junction Compensation
From above discussions, it is imminent that the thermocouple output voltage will vary if the reference junction temperature changes. So, for measurement of temperature, it is desirable that the cold junction of the thermocouple should be maintained at a constant temperature. Ice bath can be used for this purpose, but it is not practical solution for industrial situation. An alternative is to use a thermostatically controlled constant temperature oven. In this case, a fixed voltage must be added to the voltage generated by the thermocouple, to obtain the actual temperature. But the most common case is where the reference junction is placed at ambient temperature. For high temperature measurement, the error introduced due to variation of reference junction temperature is not appreciable. For example, with a thermocouple measuring 1500oC, if the ambient temperature variation is within 25 ± 15 oC, then the error introduced in the reading due to this variation would be around 1%. Such a typical scheme is shown in fig. 7. Here a constant voltage corresponding to the ambient temperature is added through the offset of the op-amp. The thermocouple voltage is also amplified by the same op-amp. A more accurate method for reference junction temperature compensation is shown in fig. 8. Here a thermistor, or a RTD is used to measure the ambient temperature and compensate the error through a bridge circuit. The bridge circuit is balanced at 0oC. When the ambient temperature goes above 0oC, the emf generated in the thermocouple is reduced; at the same time bridge unbalanced voltage is added to it in order to maintain the overall voltage at the same value. R2
R1 Hot junction Cold junction + Offset adjust

Fig.7 Simple method for reference junction compensation

Version 2 EE IIT, Kharagpur 12

Thermistor

R

RP R Hot junction Cold junction To amplifier

Fig. 8 Compensation scheme using wheatstone bridge

As referred to Fig.8, the cold junction compensation is normally kept along with the signal conditioning circuits, away from the measuring point. This may require use of long thermocouple wires to the compensation circuit. In order to reduce the length of costly thermocouple wires (platinum in some case) low-cost compensating wires are normally used in between the thermocouple and the compensation circuit. These wires are so selected that their temperature emf characteristics match closely to those of the thermocouple wires around the ambient temperature.

4.

Conclusion

Temperature is the most important process variable that requires continuous measurement and monitoring in a process industry. Among the different types of temperature transducers, the most commonly used ones are RTDs and thermocouples. Their popularity is mainly due to their ruggedness, repeatability and wide range of operation. Bare RTDs and thermocouples are rarely used in practice; instead, they are put in protective metallic sheaths. The signal conditioning circuits should be properly designed, so as to avoid the errors due to lead wires in RTDs and variation of cold junction temperatures in thermocouples. There are several cases where the temperature to be measured is more than 2000oC, the conventional measuring techniques fail to measure the high temperature. Instead, the measurement is carried out from a distance. Radiation pyrometers are used in these situations. However, its principle of operation has not been included in this lesson.

Review Exercise
1. Name the materials commonly used for RTDs. Which one has the most linear characteristics? 2. What do you mean by lead wire compensation of RTD. How can they be achieved? 3. PT-100 is a Platinum RTD whose resistance at 0oC is 100Ω. If the resistance 4. temperature co-efficient of Platinum is 3.91X 10 −3 /oC, then find its resistance at 100oC. 5. What is the difference between a NTC thermistor and a PTC thermistor? 6. A thermistor is more suitable for measurement of temperature within a small rangejustify. Version 2 EE IIT, Kharagpur 13

7. Name three types thermocouples and their suitable temperature range. 8. State and explain the laws of thermocouple. 9. What do you mean by cold junction compensation of a thermocouple? Suggest two methods for compensation. 10. A thermocouple has a linear sensitivity of 30µv/oC, calibrated at a cold junction temperature of 0oC. It is used measure an unknown temperature with the cold junction temperature of 30oC. Find the actual hot junction temperature if the emf generated is 3.0 mv.

Answers
1. Platinum, Copper, Nickel; Platinum. 3. 139 Ω . 9. 130oC.

Version 2 EE IIT, Kharagpur 14

Module 2
Measurement Systems
Version 2 EE IIT, Kharagpur 1

Lesson 5
Pressure and Force Measurement
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
The reader, after going through the lesson would be able to 1. Name different methods for pressure measurement using elastic transducers. 2. Explain the construction and principle of operation of a Bourdon tube pressure gage. 3. Define gage factor of a strain gage 4. Name different strain gage materials and state their gage factors. 5. Will be able to draw the connection diagram of an unbalanced bridge with four strain gages so as to obtain maximum sensitivity and perfect temperature compensation. 6. Name different methods for force measurement with strain gages.

1.

Introduction

In this lesson, we will discuss different methods for measurement of pressure and force. Elastic elements, namely diaphragms and Bourdon tubes are mainly used for pressure measurement. On the other hand, strain gages are commonly used for measurement of force. The constructions and principles of operation of different elastic elements for pressure measurement have been discussed in the next section. This is followed by principle of strain gage and measurement of force using strain gages.

2.

Pressure Measurement

Measurement of pressure inside a pipeline or a container in an industrial environment is a challenging task, keeping in mind that pressure may be very high, or very low (vacuum); the medium may be liquid, or gaseous. We will not discuss the vacuum pressure measuring techniques; rather try to concentrate on measurement techniques of pressure higher than the atmospheric. They are mainly carried out by using elastic elements: diaphragms, bellows and Bourdon tubes. These elastic elements change their shape with applied pressure and the change of shape can be measured using suitable deflection transducers. Their basic constructions and principle of operation are explained below.

2.1 Diaphragms
Diaphragms may be of three types: Thin plate, Membrane and Corrugated diaphragm. This classification is based on the applied pressure and the corresponding displacements. Thin plate (fig. 1(a)) is made by machining a solid block and making a circular cross sectional area with smaller thickness in the middle. It is used for measurement of relatively higher pressure. In a membrane the sensing section is glued in between two solid blocks as shown in fig. 1(b). The thickness is smaller; as a result, when pressure is applied on one side, the displacement is larger. The sensitivity can be further enhanced in a corrugated diaphragm (fig. 1(c)), and a large deflection can be obtained for a small change in pressure; however at the cost of linearity. The materials used are Bronze, Brass, and Stainless steel. In recent times, Silicon has been extensively used the diaphragm material in MEMS (Micro Electro Mechanical Systems) pressure sensor. Further, the natural frequency of a diaphragm can be expressed as: Version 2 EE IIT, Kharagpur 3

fn =

1 2π

k meq

(1)

where meq = equivalent mass, and k= elastic constant of the diaphragm. The operating frequency of the pressure to be measured must be less than the natural frequency of the diaphragm.

When pressure is applied to a diaphragm, it deflates and the maximum deflection at the centre ( y0 ) can measured using a displacement transducer. For a Thin plate, the maximum deflection

y0 is small ( y0 < 0.3 t ) and referring fig. 2, a linear relationship between p and y0 exists as:
3 (1 − ν 2 ) 4 p R 16 Et 3 where, E= Modulus of elasticity of the diaphragm material, and ν = Poisson’s ratio. However, the allowable pressure should be less than: y0 =

(2)

⎛t ⎞ pmax = 1.5 ⎜ ⎟ σ max ⎝R⎠ where, σ max is the safe allowable stress of the material.

2

(3)

For a membrane, the deflection is larger, and the relationship between p and y0 is nonlinear and can be expressed as (for ν = 0.3): E t3 3 (4) p = 3.58 4 y0 R

Version 2 EE IIT, Kharagpur 4

For a corrugated diaphragm, it is difficult to give any definite mathematical relationship between p and y0 ; but the relationship is also highly nonlinear. As the diaphragm deflates, strains of different magnitudes and signs are generated at different locations of the diaphragm. These strains can also be measured by effectively placing four strain gages on the diaphragm. The principle of strain gage will be discussed in the next section.

2.2

Bellows

Bellows (fig. 3) are made with a number of convolutions from a soft material and one end of it is fixed, wherein air can go through a port. The other end of the bellows is free to move. The displacement of the free end increases with the number of convolutions used. Number of convolutions varies between 5 to 20. Often an external spring is used opposing the movement of the bellows; as a result a linear relationship can be obtained from the equation: pA=kx (5) where, A is the area of the bellows, k is the spring constant and x is the displacement of the bellows. Phosphor Bronze, Brass, Beryllium Copper, Stainless Steel are normally used as the materials for bellows. Bellows are manufactured either by (i) turning from a solid block of metal, or (ii) soldering or welding stamped annular rings, or (iii) rolling (pressing) a tube.

2.3

Bourdon Tube

Bourdon tube pressure gages are extensively used for local indication. This type of pressure gages were first developed by E. Bourdon in 1849. Bourdon tube pressure gages can be used to measure over a wide range of pressure: form vacuum to pressure as high as few thousand psi. It is basically consisted of a C-shaped hollow tube, whose one end is fixed and connected to the pressure tapping, the other end free, as shown in fig. 4. The cross section of the tube is elliptical. When pressure is applied, the elliptical tube tries to acquire a circular cross section; as a result, stress is developed and the tube tries to straighten up. Thus the free end of the tube moves up, depending on magnitude of pressure. A deflecting and indicating mechanism is attached to the free end that rotates the pointer. The materials used are commonly Phosphor Bronze, Brass and Beryllium Copper. For a 2" overall diameter of the C-tube the useful travel of the free end is Version 2 EE IIT, Kharagpur 5

1 . Though the C-type tubes are most common, other shapes of tubes, such as 8 helical, twisted or spiral tubes are also in use.
approximately

"

3.

Measurement of Force

The most popular method for measuring force is using strain gage. We measure the strain developed due to force using strain gages; and by multiplying the strain with the effective cross sectional area and Young’s modulus of the material, we can obtain force. Load cells and Proving rings are two common methods for force measurement using strain gages. We will first discuss the principle of strain gage and then go for the force measuring techniques.

3.1 Strain Gage
Strain gage is one of the most popular types of transducer. It has got a wide range of applications. It can be used for measurement of force, torque, pressure, acceleration and many other parameters. The basic principle of operation of a strain gage is simple: when strain is applied to a thin metallic wire, its dimension changes, thus changing the resistance of the wire. Let us first investigate what are the factors, responsible for the change in resistance.

3.1.1 Gage Factor
Let us consider a long straight metallic wire of length l circular cross section with diameter d (fig. 5). When this wire is subjected to a force applied at the two ends, a strain will be generated and as a result, the dimension will change (l changing to l + Δl , d changing to d + Δd and A changing to A + ΔA ). For the time being, we are considering that all the changes are in positive direction. Now the resistance of the wire:
Version 2 EE IIT, Kharagpur 6

, where ρ is the resistivity. A From the above expression, the change in resistance due to strain: ⎛ ∂R ⎞ ⎛ ∂R ⎞ ⎛ ∂R ⎞ ΔR = ⎜ ⎟ Δl + ⎜ ⎟ ΔA + ⎜ ⎜ ∂ρ ⎟ Δρ ⎟ ⎝ ∂A ⎠ ⎝ ∂l ⎠ ⎝ ⎠ ρ ρ l = Δl − 2 ΔA + Δρ A A A Δl ΔA Δρ =R −R +R ρ A l or, ΔR Δl ΔA Δρ (6) = − + ρ R l A
R=

ρl

Now, for a circular cross section, A =
ΔA Δd =2 A d
Hence,

πd 2
4

; from which, ΔA =

πd
2

Δd . Alternatively,

ΔR Δl Δd Δ ρ (7) = −2 + ρ R l d Now, the Poisson’s Ratio is defined as: Δd lateral strain d υ =− =− Δl longitudinal strain l The Poisson’s Ratio is the property of the material, and does not depend on the dimension. So, (6) can be rewritten as: ΔR Δl Δρ = (1 + 2υ ) + ρ R l Hence, Version 2 EE IIT, Kharagpur 7

ΔR

ρ R = 1 + 2υ + Δl Δl l l

Δρ

The last term in the right hand side of the above expression, represents the change in resistivity of the material due to applied strain that occurs due to the piezo-resistance property of the material. In fact, all the elements in the right hand side of the above equation are independent of the geometry of the wire, subjected to strain, but rather depend on the material property of the wire. Due to this reason, a term Gage Factor is used to characterize the performance of a strain gage. The Gage Factor is defined as: Δρ ΔR ρ R = 1 + 2υ + G := (8) Δl Δl l l For normal metals the Poisson’s ratio υ varies in the range: 0.3 ≤ υ ≤ 0.6 , while the piezo-resistance coefficient varies in the range: Δρ 0.2 ≤

ρ

l Thus, the Gage Factor of metallic strain gages varies in the range 1.8 to 2.6. However, the semiconductor type strain gages have a very large Gage Factor, in the range of 100-150. This is attained due to dominant piezo-resistance property of semiconductors. The commercially available strain gages have certain fixed resistance values, such as, 120Ω, 350 Ω, 1000 Ω, etc. The manufacturer also specifies the Gage Factor and the maximum gage current to avoid selfheating (normally in the range 15 mA to 100 mA). The choice of material for a metallic strain gage should depend on several factors. The material should have low temperature coefficient of resistance. It should also have low coefficient for thermal expansion. Judging from all these factors, only few alloys qualify for a commercial metallic strain gage. They are: Advance (55% Cu, 45% Ni): Gage Factor between 2.0 to 2.2 Nichrome (80% Ni, 20% Co): Gage Factor between 2.2 to 2.5 Apart from these two, Isoelastic -another trademarked alloy with Gage Factor around 3.5 is also in use. Semiconductor type strain gages, though having large Gage Factor, find limited use, because of their high sensitivity and nonlinear characteristics.

Δl

≤ 0.6 .

Version 2 EE IIT, Kharagpur 8

3.1.2 Metallic Strain Gage
Most of the strain gages are metallic type. They can be of two types: unbonded and bonded. The unbonded strain gage is normally used for measuring strain (or displacement) between a fixed and a moving structure by fixing four metallic wires in such a way, so that two are in compression and two are in tension, as shown in fig. 6 (a). On the other hand, in the bonded strain gage, the element is fixed on a backing material, which is permanently fixed over a structure, whose strain has to be measured, with adhesive. Most commonly used bonded strain gages are metal foil type. The construction of such a strain gage is shown in fig. 6(b). The metal foil type strain gage is manufactured by photo-etching technique. Here the thin strips of the foil are the active elements of the strain gage, while the thick ones are for providing electrical connections. Because of large area of the thick portion, their resistance is small and they do not contribute to any change in resistance due to strain, but increase the heat dissipation area. Also it is easier to connect the lead wires with the strain gage. The strain gage in fig. 6(b) can measure strain in one direction only. But if we want to measure the strain in two or more directions at the same point, strain gage rosette, which is manufactured by stacking multiple strain gages in different directions, is used. Fig. 7 shows a three-element strain gage rosette stacked at 450.

Version 2 EE IIT, Kharagpur 9

The backing material, over which the strain gage is fabricated and which is fixed with the strain measuring structure has to satisfy several important properties. Firstly, it should have high mechanical strength; it should also have high dielectric strength. But the most important it should have is that it should be non-hygroscopic, otherwise, absorption of moisture will cause bulging and generate local strain. The backing materials normally used are impregnated paper, fibre glass, etc. The bonding material used for fixing the strain gage permanently to the structure should also be non-hygroscopic. Epoxy and Cellulose are the bonding materials normally used.

3.1.3 Semiconductor type Strain Gage
Semiconductor type strain gage is made of a thin wire of silicon, typically 0.005 inch to 0.0005 inch, and length 0.05 inch to 0.5 inch. They can be of two types: p-type and n-type. In the former the resistance increases with positive strain, while, in the later the resistance decreases with temperature. The construction and the typical characteristics of a semiconductor strain gage are shown in fig.8. MEMS pressure sensors is now a days becoming increasingly popular for measurement of pressure. It is made of a small silicon diagram with four piezo-resistive strain gages mounted on it. It has an in-built signal conditioning circuits and delivers measurable output voltage corresponding to the pressure applied. Low weight and small size of the sensor make it suitable for measurement of pressure in specific applications.

Version 2 EE IIT, Kharagpur 10

3.1.4 Strain Gage Bridge
Normal strain experienced by a strain gage is in the range of micro strain (typical value: 100 x 10-6). As a result, the change in resistance associated with it is small ( ΔR R = Gε ). So if a single strain gage is connected to a wheatstone bridge, with three fixed resistances, the bridge output voltage is going to be linear (recall, that we say the bridge output voltage would be linearly varying with ΔR R , if ΔR R does not exceed 0.1). But still then, a single strain gage is normally never used in a wheatstone bridge. This is not because of improving linearity, but for obtaining perfect temperature compensation. Suppose one strain gage is connected to a bridge with three fixed arms. Due to temperature rise, the strain gage resistance will change, thus making the bridge unbalance, thus giving an erroneous signal, even if no strain is applied. If two identical strain gages are fixed to the same structure, one measuring compressional strain and the other tensile strain, and connected in the adjacent arms of the bridge, temperature compensation can be achieved. If the temperature increases, both the strain gage resistances will be affected in the same way, thus maintaining the bridge balance under no strain condition. One more advantage of using the push-pull configuration is increasing the sensitivity. In fact, all the four arms of the bridge can be formed by four active gages; this will improve the sensitivity further, while retaining the temperature compensation property. A typical strain gage bridge is shown in fig. 9. It can be shown that if nominal resistances of the strain gages are same and also equal gage factor G, then the unbalanced voltage is given be:

EG ( ε1 + ε 3 − ε 2 − ε 4 ) 4 where ε1 , ε 2 , ε 3 , ε 4 are the strains developed with appropriate signs. e0 =

(9)

Version 2 EE IIT, Kharagpur 11

3.2

Load Cell

Load cells are extensively used for measurement of force; weigh bridge is one of the most common applications of load cell. Here two strain gages are fixed so as to measure the longitudinal strain, while two other measuring the transverse strain, as shown in fig. 10. The strain gages, measuring the similar strain (say, tensile) are placed in the opposite arms, while the adjacent arms in the bridge should measure opposite strains (one tensile, the other compressional). If the strain gages are identical in characteristics, this will provide not only the perfect temperature coefficient, but also maximum obtainable sensitivity from the bridge. The longitudinal strain developed in the load cell would be compressional in nature, and is given F , where F is the force applied, A is the cross sectional area and Y is the Young’s by: ε1 = − A E modulus of elasticity. The strain gages 1 and 3 will experience this strain, while for 2 and 4 the νF strain will be ε 2 = , where ν is the Poisson’s ratio. AE

Version 2 EE IIT, Kharagpur 12

3.3 Proving Ring
Proving Rings can be used for measurement of both compressional and tensile forces. The advantage of a Proving Ring is that, because of its construction more strain can be developed compared to a load cell. The typical construction of a Proving Ring is shown in fig.11. It consists of a hollow cylindrical beam of radius R, thickness t and axial width b. The two ends of the ring are fixed with the structures between which force is measured. Four strain gages are mounted on the walls of the proving ring, two on the inner wall, and two on the outer wall. When force is applied as shown, gages 2 and 4 will experience strain –ε (compression), while gages 1 and 3 will experience strain + ε (tension). The magnitude of the strain is given by the expression: 1.08FR (10) ε= Ebt 2 The four strain gages are connected in a bridge and the unbalanced voltage can easily be calibrated in terms of force to be measured.

3.4 Cantilever Beam
Cantilever beam can be used for measurement up to 10 kg of weight. One end of the cantilever is fixed, while the other end is free; load is applied at this end, as shown in fig. 12. The strain developed at the fixed end is given by the expression: 6Fl ε= (11) Ebt 2 where, l = length of the beam t = thickness of the cantilever b = width of the beam E = Young’s modulus of the material The strain developed can be measured by fixing strain gages at the fixed end: two on the top side of the beam, measuring tensile strain +ε and two on the bottom measuring compressional strain – ε (as shown in fig. 12) and using eqn. (9).

Version 2 EE IIT, Kharagpur 13

4.

Conclusion

In this lesson, we have studied the commonly used sensing elements for measurement of pressure and force. Elastic elements are used for measurement of pressure, where the pressure signal is converted into displacement signal. Displacement sensors are further used to convert this to appropriate electrical signal. Strain gages are also sometimes used to measure strain developed on the diaphragm. On the other hand, load cells, Proving Rings and Cantilever Beams are used for force measurement. Here strain gages mounted on the sensing elements measure strains, and the unbalanced voltage of a strain gage bridge can be effectively calibrated in terms of force. Another method of force measurement is using magnetostrictive transducers; but its principle of operation is beyond the scope of this lesson.

Review Exercise
1. Which one of the elastic transducers: Bellows, Thin Plate and Corrugated Diaphragm, can be used for measurement of high pressure? 2. Bellows are commonly used in conjunction with a spring. Why? 3. Explain the construction and principle of operation of a Bourdon tube pressure gage. 4. Define gage factor of strain gage. What are the strain gage materials normally used? Which one of them is having maximum gage factor? 5. What is a strain gage rosette? 6. A 120 Ω strain gage of Gage Factor 2.0 is subjected to a positive strain of 1× 10−6 . Find the change in resistance. 7. How the effect of temperature variation can be compensated in a strain gage bridge? Version 2 EE IIT, Kharagpur 14

8. How would you connect four strain gages on a cantilever beam so as to achieve maximum sensitivity and perfect temperature compensation? Show the arrangement of placing the strain gages and the bridge arrangement.

Answer
Q6. 0.24 mΩ (increase).

Version 2 EE IIT, Kharagpur 15

Module 2
Measurement Systems
Version 2 EE IIT, Kharagpur 1

Lesson 7
Flow Measurement
Version 2 EE IIT, Kharagpur 2

Instructional Objective
The reader, after going through the lesson would be able to: 1. Name different types of flowmeters, frequently used in industry. 2. Distinguish the constructional differences between orifice meter and ventury meter. 3. Understand the basic principle of operation of an obstruction type flowmeter. 4. Explain the basic principles of operation of turbine type flowmeter and electromagnetic flowmeter. 5. Develop a schematic block diagram for signal conditioning circuit for a typical flowmeter.

1. Introduction
Accurate measurement of flow rate of liquids and gases is an essential requirement for maintaining the quality of industrial processes. In fact, most of the industrial control loops control the flow rates of incoming liquids or gases in order to achieve the control objective. As a result, accurate measurement of flow rate is very important. Needless to say that there could be diverse requirements of flow measurement, depending upon the situation. It could be volumetric or mass flow rate, the medium could be gas or liquid, the measurement could be intrusive or nonintrusive, and so on. As a result there are different types of flow measuring techniques that are used in industries. The common types of flowmeters that find industrial applications can be listed as below: (a) Obstruction type (differential pressure or variable area) (b) Inferential (turbine type), (c) Electromagnetic, (d) Positive displacement (integrating), (e) fluid dynamic (vortex shedding), (f) Anemometer, (g) ultrasonic and (h) Mass flowmeter (Coriolis). In this lesson, we would learn about the construction and principle of operation few types of flowmeters.

2. Obstruction type flowmeter
Obstruction or head type flowmeters are of two types: differential pressure type and variable area type. Orifice meter, Venturimeter, Pitot tube fall under the first category, while rotameter is of the second category. In all the cases, an obstruction is created in the flow passage and the pressure drop across the obstruction is related with the flow rate.

Basic Principle
It is well know that flow can be of two types: viscous and turbulent. Whether a flow is viscous or turbulent can be decided by the Reynold’s number RD. If RD > 2000, the flow is turbulent. In the present case we will assume that the flow is turbulent, that is the normal case for practical situations. We consider the fluid flow through a closed channel of variable cross section, as Version 2 EE IIT, Kharagpur 3

shown in fig. 1. The channel is of varying cross section and we consider two cross sections of the channel, 1 and 2. Let the pressure, velocity, cross sectional area and height above the datum be expressed as p1, v1, A1 and z1 for section 1 and the corresponding values for section 2 be p2, v2, A2 and z2 respectively. We also assume that the fluid flowing is incompressible. Now from Bernloulli’s equation: p1 v12 p v2 + + z1 = 2 + 2 + z2 (1) γ 2g γ 2g where γ is the specific weight of the fluid. 2 1 Flow

p1 v1 z1

p2 v2 z2

Fig. 1 Flow through a varying cross section

If z1=z2, then

γ

p1

+

v12 p v2 = 2+ 2 2g γ 2g

(2)

If the fluid is incompressible, then v1 A1 = v2 A2 . Therefore, 2g 2 v2 − v12 = ( p1 − p2 ) γ or, A2 2g 2 ( p1 − p2 ) v2 (1 − 2 ) = 2 A1 γ Therefore, 1 2g 1 2g v2 = ( p1 − p2 ) = ( p1 − p2 ) 2 4 γ γ A2 1− β (1 − 2 ) A1 Considering circular cross section, we define β as the ratio of the two diameters, i.e. d A β = 2 , and so, 2 = β 2 . d1 A1 Therefore, the volumetric flow rate through the channel can be expressed as:

Version 2 EE IIT, Kharagpur 4

Q = v2 A2 =

A2 1− β
4

2g

γ

( p1 − p2 )

(3)

From the above expression, we can infer that if there is an obstruction in the flow path that causes the variation of the cross sectional area inside the closed flow channel, there would be difference in static pressures at two points and by measuring the pressure difference, one can obtain the flow rate using eqn. (3). However, this expression is valid for incompressible fluids (i.e. liquids) only and the relationship between the volumetric flow rate and pressure difference is nonlinear. A special signal conditioning circuit, called square rooting circuit is to be used for getting a linear relationship.

Orifice meter
Depending on the type of obstruction, we can have different types of flow meters. Most common among them is the orifice type flowmeter, where an orifice plate is placed in the pipe line, as shown in fig.2. If d1 and d2 are the diameters of the pipe line and the orifice opening, then the flow rate can be obtained using eqn. (3) by measuring the pressure difference (p1-p2).
Flow profile Orifice Plate Vena Contacta

Flow d1 d2

p1

p2 Fig. 2 Orifice type flowmeter

Corrections
The flow expression obtained from eqn.(3) is not an accurate expression in the actual case, and some correction factor, named as discharge co-efficient (Cd) has to be incorporated in (3), as C A 2g ( p1 − p2 ) Q = v2 A2 = d 2 (4) 4 γ 1− β Cd is defined as the ratio of the actual flow and the ideal flow and is always less than one. There are in fact two main reasons due to which the actual flow rate is less than the ideal one (obtained from eqn. (3)). The first is that the assumption of frictionless flow is not always valid. The amount of friction depends on the Reynold’s number (RD). The more important point is that, the minimum flow area is not the orifice area A2, but is somewhat less and it occurs at a distance from the orifice plate, known as the Vena Contracta, and we are taking a pressure tapping around Version 2 EE IIT, Kharagpur 5

that point in order to obtain the maximum pressure drop. As a result, the correction factor Cd <1, has to be incorporated. In fact Cd depends on β, as well as on RD. But it has been observed that for RD>104, the flow is totally turbulent and Cd is independent on RD. In this range, the typical value of Cd for orifice plate varies between 0.6 and 0.7.

Orifice Plate, Venturimeter and Flow nozzle
The major advantages of orifice plate are that it is low cost device, simple in construction and easy to install in the pipeline as shown in fig.3. The orifice plate is a circular plate with a hole in the center. Pressure tappings are normally taken distances D and 0.5D upstream and downstream the orifice respectively (D is the internal diameter of the pipe). But there are many more types of pressure tappings those are in use.

Permanent Pressure drop

Fig. 3 Orifice plate and permanent pressure drop

The major disadvantage of using orifice plate is the permanent pressure drop that is normally experienced in the orifice plate as shown in fig.3. The pressure drops significantly after the orifice and can be recovered only partially. The magnitude of the permanent pressure drop is around 40%, which is sometimes objectionable. It requires more pressure to pump the liquid. This problem can be overcome by improving the design of the restrictions. Venturimeters and flow nozzles are two such devices. The construction of a venturimeter is shown in fig.4. Here it is so designed that the change in the flow path is gradual. As a result, there is no permanent pressure drop in the flow path. The discharge coefficient Cd varies between 0.95 and 0.98. The construction also provides high mechanical strength for the meter. However, the major disadvantage is the high cost of the meter.

Version 2 EE IIT, Kharagpur 6

Flow nozzle is a compromise between orifice plate and venturimeter. The typical construction is shown in fig. 5.

p1

p2 Fig. 4 Venturimeter

p1 Fig. 5 Flow nozzle

p2

In general, few guidelines are to be followed for installation of obstruction type flowmeters. Most important among them is that, no other obstruction or bending of the pipe line is not allowed near the meter. Though this type of flowmeters are most popular in industries, their accuracy is low for low flow rates. As a result, they are not recommended for low flow rate measurement.

Flow measurement of compressible fluids
So far we have discussed about the flow measurement of incompressible fluids (liquids). For of compressible fluids, i.e. gases, the flow rates are normally expressed in terms of mass flow rates. The same obstruction type flowmeters can be used, but an additional correction factor needs to be introduced to take in to account the compressibility of the gas used. The mass flow rate gases can be expressed as :

⎡ C A W =Y ⎢ d 2 ⎢ 1− β 4 ⎣
where,

2 g ( p1 − p2 ) ⎤ ⎥ v1 ⎥ ⎦

(5)

v1= specific volume of the gas in m3/kgf

Version 2 EE IIT, Kharagpur 7

p1 − p2 1 p1 K C K= Specific heat ratio p of the gas at state 1 Cv And all other terms are as defined in (1). Y = 1 − (0.41 + 0.35β 4 )

Pitot Tube
Pitot tube is widely used for velocity measurement in aircraft. Its basic principle can be understood from fig. 6(a). If a blunt object is placed in the flow channel, the velocity of fluid at the point just before it, will be zero. Then considering the fluid to be incompressible, from eqn. (2), we have, p1 v12 p v2 + = 2+ 2 γ 2g γ 2g Now v 2 = 0 . Therefore,
2 v1 p − p1 = 2 2g γ

or,

v1 =

2g

γ

( p 2 − p1 )

(6)

However, as mentioned earlier corrections are to be incorporated for compressible fluids. The typical construction of a Pitot tube is shown in fig. 6(b).
Blunt object

PL V1

V2 = 0 Fig. 6(a) Pitot Tube: Basic Principle

p1

p2

Fig. 6(b) Pitot Tube: Construction

Rotameter
The orificemeter, Venturimeter and flow nozzle work on the principle of constant area variable pressure drop. Here the area of obstruction is constant, and the pressure drop changes with flow rate. On the other hand Rotameter works as a constant pressure drop variable area meter. It can be only be used in a vertical pipeline. Its accuracy is also less (2%) compared to other types of flow meters. But the major advantages of rotameter are, it is simple in construction, ready to install and the flow rate can be directly seen on a calibrated scale, without the help of any other device, e.g. differential pressure sensor etc. Moreover, it is useful for a wide range of variation of flow rates (10:1). Version 2 EE IIT, Kharagpur 8

The basic construction of a rotameter is shown in fig. 7. It consists of a vertical pipe, tapered downward. The flow passes from the bottom to the top. There is cylindrical type metallic float inside the tube. The fluid flows upward through the gap between the tube and the float. As the float moves up or down there is a change in the gap, as a result changing the area of the orifice. In fact, the float settles down at a position, where the pressure drop across the orifice will create an upward thrust that will balance the downward force due to the gravity. The position of the float is calibrated with the flow rate.
Float p2 Tapered pipe

p1 Orifice area

Flow Fig. 7 Basic construction of a rotameter.

Let us consider,

γ 1 = Specific weight of the float γ 2 = specific weight of the fluid v f = volume of the float
Af = Area of the float.

At = Area of the tube at equilibrium (corresponding to the dotted line) From equation (4), for incompressible fluid, we have, for the orifice, Cd A2 2g ( p1 − p2 ) Q= (7) A2 2 γ 2 1− ( ) A1 Now consider the free body diagram of the float, shown in fig. 8. Let,
Fd W

Fu Fig. 8 Forces acting on the float

Fd = Downward thrust on the float Fu = Upward thrust on the float
Version 2 EE IIT, Kharagpur 9

W = Apparent weight of the float At balance, W = Fu − Fd

V f (γ 1 − γ 2 ) = p1 Af − p2 Af or, Therefore, vf p1 − p2 = (γ 1 − γ 2 ) Af Substituting the above expression in (7), we obtain:
Q= ⎡ 2g v f ⎤ (γ 1 − γ 2 ) ⎥ ⎢ 2 ⎥ ⎧ At − Af ⎫ ⎢ γ 2 Af ⎣ ⎦ 1− ⎨ ⎬ ⎩ At ⎭ Cd ( At − Af )

(8)

⎧ At − Af ⎫ The term within the third bracket in the above expression is constant. If ⎨ ⎬ << 1 , then, ⎩ At ⎭ we can have, Q = K ( At − Af ) If the tube is made in such a way that At varies linearly with the displacement, one have a linear relationship in the form, (9) Q = K1 + K 2 x that is, the scale of the tube can be graduated linearly in terms of flow rate. Otherwise, the displacement of the float can be converted to electrical signal by using a LVDT or similar type of displacement sensor. For large flow rate measurement, the rotameter is normally place in a bypass line.
The major source of error in rotameter is due to the variation of density of the fluid. Besides, the presence of viscous force may also provide an additional force to the float.

2

Construction of the float
The construction of the float decides heavily, the performance of the rotameter. In general, a float should be designed such that: (a) it must be held vertical (b) it should create uniform turbulence so as to make it insensitive to viscosity (c) it should make the rotameter least sensitive to the variation of the fluid density. A typical construction of the float is shown in fig. 9. The top section of the float has a sharp edge and several angular grooves. The fluid passing through these grooves, causes the rotation of the float. The turbulence created in this process reduces the viscous force considerably.

Version 2 EE IIT, Kharagpur 10

Angular grooves Sharp edge

Fig. 9 Construction of a float

From (8) the expression for volumetric flowrate can be written as: 2g v f Q=K (γ − γ ) (10) γ 2 Af 1 2 The performance of the flowmeter can be made almost independent of the variation of fluid density, if we select the material of the float, such that, γ 1 >> γ 2 . For measurement of mass flow rate (W), we can write, W = γ 2Q = K K1 (γ 1 − γ 2 )γ 2 (11) dW = 0 , can be satisfied, if we select γ 1 = γ 2 . This can be achieved by using a The condition, dγ 2 hollow float, or a plastic float.

Electromagnetic Flowmeter
Electromagnetic flowmeter is different from all other flowmeters due to its uniqueness on several accounts. The advantages of this type of flowmeter can be summarized as: 1. It causes no obstruction to flow path. 2. It gives complete linear output in form of voltage. 3. The output is unaffected by changes in pressure, temperature and viscosity of the fluid. 4. Reverse flow can also be measured. 5. Flow velocity as low as 10-6m/sec can be measured.

Electrodes e0 v

B v

B e0

Fig. 10 Electromagnetic Flowmeter

Version 2 EE IIT, Kharagpur 11

Electromagnetic flowmeters are suitable for measurement of velocity of conducting (Mercury) and weakly conducting (water) liquid. The basic principle of operation can be understood from fig. 10. It works on the principle of basic electromagnetic induction; i.e. when a conductor moves along a magnetic field perpendicular to the direction of flow, a voltage would be induced perpendicular to the direction of movement as also to the magnetic filed. The flowing liquid acts like a conductor. External magnetic field is applied perpendicular to the direction of the flow and two electrodes are flushed on the wall of the pipeline as shown. The expression for the voltage induced is given by: eo = B l v (12) where l is the length of the conductor (diameter d in this case) and v is the velocity of the liquid. The above expression shows the complete relationship between the voltage induced and the velocity. However, the magnetic field applied is not d.c. if the liquid medium is water or any other polarizable liquid. This is because, if the magnetic field is d.d. the voltage induced will also be d.c. and a small amount of d.c. current will flow if a measuring circuit is connected to the terminals. This small d.c. current will cause electrolysis; oxygen and hydrogen bubbles will be formed and they will stick to the electrodes surfaces for some time. This will provide an insulating layer on the electrodes surfaces that will disrupt the voltage generation process. As a result, the magnetic field applied for these cases is a.c., or pulsed d.c. excitation. The meter can only be used for liquids having moderate conductivities (more than 10 μ mho / cm ). As a result, it is not suitable for gases or liquid hydrocarbons. The accuracy is around ±1% .

Turbine type Flowmeter
Turbine type flowmeter is a simple way for measuring flow velocity. A rotating shaft with turbine type angular blades is placed inside the flow pipe. The fluid flowing through the pipeline will cause rotation of the turbine whose speed of rotation can be a measure of the flowrate. Referring fig.11, let blades make an angle α with the body. Then, ωr R = tan α − v where, − Q v = Average velocity of the fluid = A Q = Volumetric flowrate A = Effective flow area of the pipe R = Radius of the blade ω r = Angular speed of the blade. From the above expression, the volumetric flow rate can be related with the angular speed, as: (13) ωr = k Q where, tan α k= RA

Version 2 EE IIT, Kharagpur 12

ωr

ωrR

α v

Fig. 11 Turbine type flowmeter

The speed of rotation of the turbine can be measured using several ways, such as, optical method, inductive pick up etc.

Vortex type Flowmeter
Formation of vortex on a flowing stream by an obstruction like straw or stone is a common observation. But what is probably not commonly known is the fact that, the frequency of vortex formation is proportional to flow velocity.
Transmitter

d Flow

Blunt object

Receiver Karman Votex

Fig. 12 Vortex type flowmeter

Fig.12 shows the basic principle of vortex type flowmeter. It is based on the principle of vertex shading. When a blunt object is placed on the passage of a flowing stream, vortices are formed. A vortes of this sort is called Karman Vortex. If the flow is turbulent and the Reynold’s number is RD > 104 , then the frequency of vortex formation is given by: N f = st v (14) d where, d= width of the blunt object Version 2 EE IIT, Kharagpur 13

v = velocity of the fluid Nst = A constant, called Strouhal Number. The fig. 12 shows a typical arrangement of measurement of frequency of vorticex formation using ultrasonic technique. Formation of a vortex will modulate the intensity of ultrasound received by the receiver, and the frequency of modulation can be measured easily.

Conclusion
In this lesson, we have learnt about various techniques of flow measurement in industrial processes. It has been seen that most of the flow measurement techniques are based on the principle of obstruction type flowmeter. Orifice meters and venture meters are the two most popular types of transducers for flow measurement. However, they require, additional differential pressure transducers for converting the differential pressure generated into appropriate electrical signals and also square rooting devices in order to obtain a linear output proportional to flowrate. Comparatively, electromagnetic flowmeter provides a direct method for measurement of flowrate and gives a proportional voltage output with respect to flow. It also does not provide any obstruction to the flow path; as a result, there is no pressure drop. But this technique is suitable for conducting fluids only and cannot be used for gases. Moreover, often the polarization property of water creates problems and calls for an involved signal conditioning circuit. There are few other types of flowmeters, whose principles of operations could not be discussed here due to paucity of space. One of them is the ultrasonic flowmeter. This type of flowmeter is also non-intrusive type, i.e., it does not provide any obstruction to the flow passage. But it is quite costly, compared to other flowmeters. Positive displacement flowmeter is an integral type of flowmeter, in the sense, that it measures total flow in a given amount of time, and finds wide use in water meters, petrol pumps etc. Its construction is normally different from other types of flowmeters, though turbine type flowmeter with a counter to count the number of revolutions can also be used for this purpose.

Review Exercise
1. 2. 3. 4. 5. 6. 7. What is meant by discharge coefficient in an orifice type flowmeter? Compare the advantages and disadvantages of an orifice meter and a venturimeter. Can a rotameter be used in a horizontal pipe line? If not, explain why? The magnetic field applied to an electromagnetic flowmeter is not constant, but time varying. Why? What are the flowmeters where the output is frequency varying with flow velocity? What is the difference between a constant area variable pressure drop flowmeter and a constant pressure drop variable area flowmeter? The pressure drop across an orifice is measured for a particular flow rate. If the flow rate is doubled, keeping all other parameters constant, what would happen to the pressure drop? a) It will remain the same. b) It will also be doubled. c) It will be halved. d) It will increase four times. Version 2 EE IIT, Kharagpur 14

8.

A rotameter designed to measure the flow rate of water is used to measure the flow rate of brine (specific gravity 1.15), without altering the scale. Would it more, or less? Justify.

Answer
Q5. Q7. Q8. Turbine type flowmeter and Vortex type flowmeter. (d) Less (refer eqn.(8)).

Version 2 EE IIT, Kharagpur 15

Module 2
Measurement Systems
Version 2 EE IIT, Kharagpur 1

Lesson 8
Measurement of Level, Humidity and pH
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
At the end of this lesson, the student will be able to: 1. Name different methods for level and moisture measurements 2. Explain the basic techniques of level and humidity measurement 3. Explain the principle of pH measurement 4. Explain the necessity of using special measuring circuit for pH measurement

1.

Introduction

Level, humidity and pH are three important process parameters and their measurement find wide application in chemical and manufacturing industries. In this chapter we would provide a brief overview of the different techniques adopted for measurement of liquid level and humidity. The basic principle of pH measurement and the construction of pH electrodes are explained in section 4.

2.

Level Measurement

There are several instances where we need to monitor the liquid level in vessels. In some cases the problem is simple, we need to monitor the water level of a tank; a simple float type mechanism will suffice. But in some cases, the vessel may be sealed and the liquid a combustible one; as a result, the monitoring process becomes more complex. Depending upon the complexity of the situation, there are different methods for measuring the liquid level, as can be summarized as follows: (a) Float type (b) Hydrostatic differential pressure gage type (c) Capacitance type (d) Ultrasonic type (e) Radiation technique. Some of the techniques are elaborated in this section.

Hydrostatic Differential Pressure type
The hydrostatic pressure developed at the bottom of a tank is given by: p=hρ g where h is the height of the liquid level and ρ is the density of the liquid. So by putting two pressure tapings, one at the bottom and the other at the top of the tank, we can measure the differential pressure, which can be calibrated in terms of the liquid level. Such a schematic arrangement is shown in Fig. 1 . The drum level of a boiler is normally measured using this basic principle. However proper care should be taken in the measurement compensate for variation of density of water with temperature and pressure.

Version 2 EE IIT, Kharagpur 3

Capacitance type
This type of sensors are widely used for chemical and petrochemical industries; and can be used for a wide range of temperature (-40 to 200 oC) and pressure variation (25 to 60 kg/cm2). It uses a coaxial type cylinder, and the capacitance is measured between the inner rod and the outer cylinder, as shown in Fig. 2. The total capacitance between the two terminals is the sum of (i) capacitance of the insulating bushing, (ii) capacitance due to air and liquid vapour and (iii) capacitance due to the liquid. If the total capacitance measured when the tank is empty is expressed as C1, then the capacitance or the liquid level of h can be expressed as: 2πε 0 (ε 1 − ε 2 )h Ct = C1 + ln (r2 / r1 ) where, ε 1 is the relative permittivity of the liquid and ε 2 is the relative permittivity of the air and liquid vapour (≈ 1) . Hence a linear relationship can be obtained with the liquid level. The advantage of capacitance type sensor is that permittivity of the liquid is less sensitive to variation of temperature and can be easily compensated.

Version 2 EE IIT, Kharagpur 4

Ultrasonic type
Ultrasonic method can be effectively used for measurement of liquid level in a sealed tank. An ultrasonic transmitter/receiver pair is mounted at the bottom of the tank. Ultrasonic wave can pass through the liquid, but gets reflected at the liquid-air interface, as shown in Fig.3. The time taken to receive the pulse is measured, that can be related with the liquid level. For accurate measurement, variation of speed of sound with the liquid density (and temperature) should be properly compensated.

Radiation technique
Radioactive technique also finds applications in measurement of level in sealed containers. Radioactive ray gets attenuated as it passes through a medium. The intensity of the radiation as it passes a distance x through a medium is given by: I ( x) = I 0 e−α x where I0 is the incidental intensity and α is the absorption co-efficient of the medium. Thus if we measure the intensity of the radiation, knowing I0 , and α, x can be determined. There are several techniques which are in use. In one method, a float with a radioactive source inside is allowed to move along a vertical path with the liquid level. A Geiger Muller Counter is placed at the bottom of the tank along the vertical path and the intensity is measured. The basic scheme is shown in Fig. 4. The method used in a batch filling process of bottles, uses a source-detector assembly that can slide along the two sides of the bottle, as shown in Fig. 5 . As soon as the source-detector assembly passes through the liquid-air interface, there would be a large change in the signal received by the detector. Radioactive methods, though simple in principle, find limited applications, because of possible radiation hazards. However radioactive methods are routinely used for level measurement of grains and granular solids.

Version 2 EE IIT, Kharagpur 5

3.

Humidity Measurement

Humidity measurement finds wide applications in different process industries. Moisture in the atmosphere must be controlled below a certain level in many manufacturing processes, e.g., semiconductor devices, optical fibres etc. Humidity inside an incubator must be controlled at a very precision level. Textiles, papers and cereals must be dried to a standard storage condition in order to prevent the quality deterioration. The humidity can be expressed in different ways: (a) absolute humidity, (b) relative humidity and (c) dew point. Humidity can be measured in different ways. Some of the techniques are explained below.

Hygrometer
Many hygroscopic materials, such as wood, hair, paper, etc. are sensitive to humidity. Their dimensions change with humidity. The change in dimension can be measured and calibrated in terms of humidity.

Psychrometer
Psychrometric method for measurement of relative humidity is a popular method. Two bulbs are used- dry bulb and wet bulb. The wet bulb is soaked in saturated water vapour and the dry bulb is kept in the ambient condition. The temperature difference between the dry bulb and wet bulb is used to obtain the relative humidity through a psychrometric chart. The whole process can also be automated.

Dew point measurement
If a gas is cooled at constant pressure to the dew point, condensation of vapour will start. The dew point can be measured by placing a clean glass mirror in the atmosphere. The temperature of the mirror surface is controlled and reduced slowly; vapour starts condensation over the mirror. Optical method is used to detect the condensation phenomena, and the temperature of the mirror surface is measured.

Conductance/Capacitance method of measurement
Many solids absorb moisture and their values of the conductance or capacitance change with the degree of moisture absorption. Moisture content in granules changes the capacitance between two electrodes placed inside. By measuring the capacitance variation, the moisture content in the Version 2 EE IIT, Kharagpur 6

granules can be measured. Similarly, moisture content in paper and textiles change their resistance. A schematic arrangement for measurement of moisture content in paper or textiles using Resistance Bridge is shown in Fig. 6.

Infrared Technique
Water molecule present in any material absorb infrared wave at wavelengths 1.94µm, 2.95 µm and 6.2µm. The degree of absorption of infrared light at any of these wavelengths may provide a measure of moisture content in the material.

4.

Measurement of pH

pH is a measure of hydrogen ion concentration in aqueous solution. It is an important parameter to determine the quality of water. The pH value is expressed as: 1 pH = log10 C Where C is the concentration of H+ ions in a solution. In pure water, the concentration of H+ ions is 10-7 gm/ltr at 25o C. So the pH value is 1 pH = =7. log10 10−7 The advantage of using pH scale is that the activities of all strong acids and bases can be brought down to the scale of 0-14. The pH value of acidic solutions is in the range 0-7 and alkaline solutions in the range 7-14. The pH value of a solution is measured by using pH electrode. It essentially consists of a pair of electrodes: measuring and reference electrode, both dipped in the solution of unknown pH. These two electrodes essentially form two half-cells; the total potential developed is the difference between the individual electric potential developed in each half cell. While the potential developed in the reference cell is constant, the measuring cell potential is dependent on the hydrogen ion concentration of the solution and is governed by Nernst’s equation: RT E = E0 + ln(a C ) nF Where: Version 2 EE IIT, Kharagpur 7

E= e.m.f of the half cell E0= e.m.f of the half cell under saturated condition R= Gas constant (8.314 J/ 0C) T= Absolute temperature (K) N= valance of the ion F= Faraday Constant = 96493 C a= Activity co-efficient (0 ≤ a ≤ 1) ; for a very dilute solution, a → 1 C= molar concentration of ions.

Measuring Electrode
The measuring electrode is made of thin sodium ion selective glass. A potential is developed across the two surfaces of this glass bulb, when dipped in aqueous solution. This potential is sensitive to the H+ ion concentration, having a sensitivity of 59.2 mv/pH at 250C. Fig. 7 shows the basic schematic of a measuring probe. The buffer solution inside the glass bulb has a constant H+ ion concentration and provides electrical connection to the lead wire.

Reference Electrode
The basic purpose of a reference electrode is to provide continuity to the electrical circuit, since the potential across a single half cell cannot be measured. With both the measuring and reference cells dipped in the same solution, the potential is measured across the two lead wires. A reference electrode should satisfy the following basic requirements: The potential developed should be independent of H+ ion concentration. (i) The potential developed should be independent of temperature (ii) (iii) The potential developed should not change with time. Considering all these requirements, two types of reference electrodes are commonly used: (i) Calomel (Mercury-Mercurous Chloride) and (ii) Silver-Silver Chloride. The construction of a Calomel reference electrode is shown in Fig. 8. The electrical connection is maintained through the salt bridge.

Version 2 EE IIT, Kharagpur 8

Sometimes the reference and measuring electrodes are housed together, as shown in Fig. 9. This type of electrode is known as Combination Electrode. The reference electrode used in this case is Silver-Silver Chloride. The combination is dipped in the solution whose pH is to be measured and the output voltage is the difference between the e.m.f.s generated by the measuring glass electrode and the reference electrode.

Version 2 EE IIT, Kharagpur 9

Measuring scheme
The sensitivity of pH probe is around 59.2mv/pH at 250C. This sensitivity should be sufficient for measurement of voltage using ordinary electronic voltmeters. But, that is not the case; special measuring circuits are required for measurement of pH voltage. This is because of the fact that the internal resistance of the pH probe as a voltage source is very high, in the order of 108-109 Ω. This is because of the fact; the electrical path between the two lead wires is completed through the glass membrane. As a result, the input resistance for of the measuring device must be at least ten times electrode resistance of the electrode. FET-input amplifier circuits are normally used for amplifying the voltage from the pH probe. Not only that, the insulation resistance between the leads must also be very high. They are normally provided with moisture resistance insulation coating. The voltage in the pH probe is temperature dependent, as evident from Nernst equation. As a result suitable temperature compensation scheme should also be provided in the measuring scheme.

Review Questions
1. How would you measure level of a liquid inside a sealed tank? Explain with a schematic arrangement any one of the methods. 2. Name different techniques used for level measurement of a liquid. Explain the principle of operation of hydrostatic differential pressure level gage. 3. Name few instances where measurement of humidity/ moisture finds important applications in industry. 4. How the moisture content in solids can be measured? Give an example and show the schematic arrangement. Version 2 EE IIT, Kharagpur 10

5. Define pH of a solution. What is the hydrogen ion concentration of a solution if the pH of the solution is 5.0? 6. Explain with simple sketches the construction of measuring electrode and reference electrode. Why two electrodes are required for pH measurement? 7. Why temperature compensation scheme should be provided in pH measurement? 8. What special arrangements are to be provided for amplifying the voltage generated in a pH electrode? Justify.

Version 2 EE IIT, Kharagpur 11

Module 2
Measurement Systems
Version 2 EE IIT, Kharagpur 1

Lesson 9
Signal Conditioning Circuits
Version 2 EE IIT, Kharagpur 2

Instructional Objective
The reader, after going through the lesson would be able to: 1. Identify the different building blocks of a measuring system and explain the function of each block. 2. Design an unbalanced wheatstone bridge and determine its sensitivity and other parameters. 3. Able to explain the advantage of using push-pull configuration in unbalanced a.c. and d.c. bridges. 4. Define CMRR of an amplifier and explain its importance for amplifying differential signal. 5. Compare the performances of single input amplifiers (inverting and non-inverting) in terms of gain and input impedance. 6. Draw and derive the gain expression of a three-op.amp. instrumentation amplifier.

1.

Introduction

It has been mentioned in Lesson-2 that a basic measurement system consists mainly of the three blocks: sensing element, signal conditioning element and signal processing element, as shown in fig.1. The sensing element converts the non-electrical signal (e.g. temperature) into electrical signals (e.g. voltage, current, resistance, capacitance etc.). The job of the signal conditioning element is to convert the variation of electrical signal into a voltage level suitable for further processing. The next stage is the signal processing element. It takes the output of the signal conditioning element and converts into a form more suitable for presentation and other uses (display, recording, feedback control etc.). Analog-to-digital converters, linearization circuits etc. fall under the category of signal processing circuits. The success of the design of any measurement system depends heavily on the design and performance of the signal conditioning circuits. Even a costly and accurate transducer may fail to deliver good performance if the signal conditioning circuit is not designed properly. The schematic arrangement and the selection of the passive and active elements in the circuit heavily influence the overall performance of the system. Often these are decided by the electrical output characteristics of the sensing element. Nowadays, many commercial sensors often have in-built signal conditioning circuit. This arrangement can overcome the problem of incompatibility between the sensing element and the signal conditioning circuit. Output Electrical output

Input Measurand

Sensing element

Signal conditioning element

Signal processing element

Fig. 1 Elements of a measuring system.

Version 2 EE IIT, Kharagpur

3

If one looks at the different cross section of sensing elements and their signal conditioning circuits, it can be observed that the majority of them use standard blocks like bridges (A.C. and D.C.), amplifiers, filters and phase sensitive detectors for signal conditioning. In this lesson, we would concentrate mostly on bridges and amplifiers and ponder about issues on the design issues.

2.

Unbalanced D.C. Bridge

We are more familiar with balanced wheatstone bridge, compared to the unbalanced one; but the later one finds wider applications in the area of Instrumentation. To illustrate the properties of unbalanced d.c. bridge, let us consider the circuit shown in fig.2 .Here the variable resistance can be considered to be a sensor, whose resistance varies with the process parameter. The output voltage is e0 , which varies with the change of the resistance x (= ΔR / R ) . The arm ratio of the bridge is p and E is the excitation voltage.

R1 = pR e0 R4 = pR

R2 = R(1+x)

R3 = R

E Fig. 2 Unbalanced D.C. bridge. Then,
⎡ R (1 + x) R ⎤ − e0 = ⎢ ⎥E ⎣ pR + R (1 + x ) pR + R ⎦ px = E ( p + 1 + x )( p + 1)

(1)

From the above expression, several conclusions can be drawn. These are: A. e0 vs. x Characteristics is nonlinear (since x is present in the denominator as well as in the numerator). B. Maximum sensitivity of the bridge can be achieved for the arm ratio p=1. Version 2 EE IIT, Kharagpur 4

The above fact can easily be verified by differentiating e0 with respect to p and equating to zero; i.e. de0 = 0 gives, dp x( p + 1 + x )( p + 1) − px ( 2 p + 2 + x) = 0 or, p 2 = 1 + x,

i.e. p = 1 + x ≈ 1 , for small x.
C.

(2)

Nonlinearity of the bridge decreases with increase in the arm ratio p, but the sensitivity is also reduced. e This fact can be verified by plotting 0 vs. x for different p, as shown in fig. 3. E D. For unity arm ratio (p=1), and for small x, we can obtain an approximate linear relationship as, x (3) e0 = E . 4
0.25 0.2

Unbalanced voltage (eo/E)

0.15 0.1 0.05

p=1

p = 10

p = 100

0 0 0.2 0.4 0.6 0.8 1
x

1.2 1.4 1.6 1.8

2

Fig. 3 Bridge characteristics for different arms ratio.

E.

We have seen that the maximum sensitivity of the bridge is attained at the arm ratio p=1. Instead of making all the values of R1, R2, R3, R4 equal under balanced condition, it could also be achieved by selecting different values with R1= R2, R3 = R4 for x = 0. But this is not advisable, since the output impedance of the bridge will be higher in the later case. So, from the requirement of low output impedance of a signal-conditioning element, it is better to construct the basic bridge with all equal resistances. It may appear from the above discussions, that, there is no restriction on selection of the bridge excitation voltage E. Moreover, since, more the excitation voltage, more is the output voltage sensitivity, higher excitation voltage is preferred. But the
Version 2 EE IIT, Kharagpur 5

F.

restriction comes from the allowable power dissipation of resistors. If we increase E, there will be more power loss in a resistance element and if it exceeds the allowable power dissipation limit, self heating will play an important role. In this case, the temperature of the resistance element will increase, which again will change the resistance and the power loss. Sometimes, this may lead to the permanent damage of the sensor (as in case of a thermistor).

Push-pull Configuration
The characteristics of an unbalanced wheatstone bridge with single resistive element as one of the arms can greatly be improved with a push-pull arrangement of the bridge, comprising of two identical resistive elements in two adjacent arms: while the resistance of one sensor decreasing, the resistance of the other sensor is increasing by the same amount, as shown in fig.4. The unbalanced voltage can be obtained as:
⎡ R(1 + x) R⎤ − e0 = ⎢ ⎥E ⎣ R(1 + x) + R(1 − x) 2 R ⎦ ⎡1 + x 1 ⎤ =⎢ − ⎥E 2⎦ ⎣ 2 x = E 2

(4)

R1 = R(1-x)

R2 = R(1+x)

e0 R3 = R

R4 = R

E Fig. 4 Unbalanced D.C. bridge with push pull configuration of resistance sensors.

Looking at the above expression, one can immediately appreciate the advantage of using pushpull configuration. First of all, the nonlinearity in the bridge output can be eliminated completely. Secondly, the sensitivity is doubled compared to a single sensor element bridge. The same concept can also be applied to A.C. bridges with inductive or capacitive sensors. These applications are elaborated below. Version 2 EE IIT, Kharagpur 6

3.

Unbalanced A.C. Bridge with Push-pull Configuration

Figures 5(a) and (b) shows the schematic arrangements of unbalanced A.C. bridge with inductive and capacitive sensors respectively with push-pull configuration. Here, the D.C. excitation is replaced by an A.C. source and two fixed resistances of same value are kept in the two adjacent arms and the inductive (or the capacitive) sensors are so designed that if the inductance (capacitance) increases by a particular amount, that of the other one would decrease by the same amount. For fig. 5(a),
⎡ jwL(1 + x) R⎤ e0 = ⎢ − ⎥E , ⎣ jwL (1 + x) + jwL(1 − x) 2 R ⎦

where w is the angular frequency of excitation, L is the nominal value of the inductance and x = ΔL . Simplifying, we obtain, L

L(1-x)

L(1+x)

C(1-x)

C(1+x)

e0 R3 = R

e0 R3 = R

R4 = R

R4 = R

~
E, ω

~
E, ω

Fig. 5 Unbalanced A.C. bridge with push-pull configuration: (a) for inductive sensor, and (b) for capacitive sensor.

x E, 2 which again shows the linear characteristics of the bridge. e0 = For the capacitance sensor with the arrangement shown in fig. 5(b), we have:

(5)

Version 2 EE IIT, Kharagpur 7

1 ⎤ ⎡ R⎥ jwC (1 + x) ⎢ E e0 = − ⎢1 2R ⎥ + 1 jwC (1 − x) ⎥ ⎢ jwC (1 + x) ⎦ ⎣ ⎡ jwC (1 − x) R⎤ =⎢ − ⎥E ⎣ jwC (1 + x) + jwC (1 − x) 2 R ⎦ x (6) =− E 2 where x = ΔC . As expected, we would also obtain here a complete linear characteristic, C irrespective of whatever is the value of x. But here is a small difference between the performance of an inductive sensor bridge and that of a capacitance sensor bridge (equation (5) and (6)): a negative sign. This negative sign in an A.C. bridge indicates that the output voltage in fig. 4(b) will be 1800 out of phase with the input voltage E. But this cannot be detected, if we use a simple A.C. voltmeter to measure the output voltage. In fact, if the value of x were negative, there would also be a phase reversal in the output voltage, which cannot be detected, unless a special measuring device for sensing the phase is used. This type of circuit is called a Phase Sensitive Device (PSD) and is often used in conjunction with inductive and capacitive sensors. The circuit of a PSD rectifies the small A.C. voltage into a D.C. one; the polarity of the D.C. output voltage is reversed, if there is a phase reversal.

Capacitance Amplifier
Here we would present another type of circuit configuration, suitable for push-pull type capacitance sensor. The circuit can also be termed as a half bridge and a typical configuration has been shown in fig.6. Here two identical voltage sources are connected in series, with their common point grounded. This can be also achieved by using a center-tapped transformer. Two sensing capacitors C1 and C2 are connected as shown in the fig. 5 and the unbalanced current flows through an amplifier circuit with a feedback capacitor Cf . Now the current through the capacitors are: I 1 = V . jwC1 and I 2 = −V . jwC 2 Hence the unbalanced current: I = I 1 + I 2 = V . jw(C1 − C 2 ) And the voltage output of the amplifier: C − C2 I =− 1 V0 = − V (7) jwC f Cf As expected, a linear response can also be obtained by connecting a push-pull configuration of capacitance in fig.6. The gain can be adjusted by varying Cf. However, this is an ideal circuit, for a practical circuit, a high resistance has to be placed in parallel with Cf.

Version 2 EE IIT, Kharagpur 8

C1 + V

Cf

~
-

+ V

~

+ C2 Fig. 6 A capacitance amplifier.

V0

4.

Amplifiers

An Amplifier is an integral part of any signal conditioning circuit. However, there are different configurations of amplifiers, and depending of the type of the requirement, one should select the proper configuration.

Inverting and Non-inverting Amplifiers
These two types are single ended amplifiers, with one terminal of the input is grounded. From the schematics of these two popular amplifiers, shown in fig.7, the voltage gain for the inverting amplifier is: e0 R =− 2 ei R1 while the voltage gain for the noninverting amplifier is: e0 R = 1+ 2 ei R1 Apparently, both the two amplifiers are capable of delivering any desired voltage gain, provided the phase inversion in the first case is not a problem. But looking carefully into the circuits, one can easily understand, that, the input impedance of the inverting amplifier is finite and is approximately R1 , while a noninverting amplifier has an infinite input impedance. Definitely, the second amplifier will perform better, if we want that, the amplifier should not load the sensor (or a bridge circuit).

Version 2 EE IIT, Kharagpur 9

R2 R2 R1 + e0 ei R1 e0 +

Fig. 7 (a) Inverting amplifier, (b) noninverting amplifier.

Differential Amplifier
Differential amplifiers are useful for the cases, where both the input terminals are floating. These amplifiers find wide applications in instrumentation. A typical differential amplifier with single op.amp. configuration is shown in fig.8. Here, by applying superposition theorem, one can easily obtain the contribution of each input and add them algebraically to obtain the output voltage as: R4 R R e0 = (1 + 2 ) e2 − 2 e1 (8) R1 R3 + R4 R1 If we select R4 R2 = , (9) R3 R1 then, the output voltage becomes: R e0 = 2 (e2 − e1 ) (10) R1
R2 R1 e1 R3 e2 + e0

R4

Fig. 8 Differential amplifier.

However, this type of differential amplifier with single op. amp. configuration also suffers from the limitation of finite input impedance. In fact, several criteria are used for judging the Version 2 EE IIT, Kharagpur 10

performance of an amplifier. These are mainly: (i) offset and drift, (ii) input impedance, (iii) gain and bandwidth, and (iv) common mode rejection ratio (CMRR). The performance of an operational amplifier is judged by the gain- bandwidth product, which is fixed by the manufacturer’s specification. In the open loop, the gain is very high (around 105) but the bandwidth is very low. In the closed loop operation, the gain is low, but the achievable bandwidth is high. Normally, the gain of a single stage operational amplifier circuit is kept limited around 10, thus large bandwidth is achievable. For larger gains, several stages of amplifiers are connected in cascade. CMRR is a very important parameter for instrumentation circuit applications and it is desirable to use amplifiers of high CMRR when connected to instrumentation circuits. The CMRR is defined as: A CMRR = 20 log10 d (11) Ac where, Ad is the differential mode gain and Ac is the common mode gain of the amplifier. The importance of using a high CMRR amplifier can be explained with the following example:

Example -1
The unbalanced voltage of a resistance bridge is to be amplified 200 times using a differential amplifier. The configuration is shown in fig. 9 with R= 1000Ω and x=2 x 10-3. Two amplifiers are available: one with Ad =200 and CMRR= 80 dB and the other with Ad =200 and CMRR= 60dB. Find the values of V0 for both the cases and compute errors.

R +10V

R2 = R(1+x) v0

Amplifier R R

Fig. 9 Solution

Here x=2 x 10-3. Using (3), x ei = x10 = 5 mv = v d 4 The common mode voltage to the amplifier is vc = 5V , half the supply voltage. Version 2 EE IIT, Kharagpur 11

For amp.-1, Ad = 200 , 20 log Therefore,

Ad = 80dB Ac

Ad 200 = 10 4 , or, Ac = 4 = 0.02 . Ac 10

v0 = Ad v d + Ac vc = 200 × 5 × 10 −3 + 0.02 × 5 = 1.1 V So, Ideally, the voltage should have been 1.0 V, 200 times the bridge unbalanced voltage, but due to the presence of common voltage, 10% error is introduced. In the second case, CMRR is 60 dB, all other values remaining same. For this case, 200 Ac = 3 = 0.2 . Therefore, 10 v0 = Ad v d + Ac vc = 200 × 5 × 10 −3 + 0.2 × 5 = 2.0 V an error of magnitude 100% is introduced due to the common mode voltage!
Referring to fig. 8, if we consider, the op. amp. to be an ideal one, then by selecting the resistances, such that, R4 R2 = , R3 R1 the effect of the common mode voltage can be eliminated completely, as is evident from eqn. (10). But if the resistance values differ, due to the tolerance of the resistors, the common mode voltage will cause error in the output voltage. The other alternative in the above example is to apply +5 and –5V at the bridge supply terminals, instead of +10V and 0V.

Instrumentation Amplifier
Often we need to amplify a small differential voltage few hundred times in instrumentation applications. A single stage differential amplifier, shown in fig.8 is not capable of performing this job efficiently, because of several reasons. First of all, the input impedance is finite; moreover, the achievable gain in this single stage amplifier is also limited due to gain bandwidth product limitation as well as limitations due to offset current of the op. amp. Naturally, we need to seek for an improved version of this amplifier. A three op. amp. Instrumentation amplifier, shown in fig.10 is an ideal choice for achieving the objective. The major properties are (i) high differential gain (adjustable up to 1000) (ii) infinite input impedance, (iii) large CMRR (80 dB or more), and (iv) moderate bandwidth. From fig. 10, it is apparent that, no current will be drown by the input stage of the op. amps. (since inputs are fed to the non inverting input terminals). Thus the second property mentioned above is achieved. Looking at the input stage, the same current I will flow through the resistances R1 and R2. Using the properties of ideal op. amp., we can have: e −e e −e e − e2 I = 1 i1 = i1 i 2 = i 2 (12) R1 R2 R1 from which, we obtain,

Version 2 EE IIT, Kharagpur 12

e1 = ei1 + e2 = ei 2 −
Therefore,

R1 (ei1 − ei 2 ) R2 R1 (ei1 − ei 2 ) R2

2 R1 )(ei1 − ei 2 ) R2 The second stage of the instrumentation amplifier is a simple differential amplifier, and hence, using (10), the over all gain: R R 2R e0 = 4 (e2 − e1 ) = 4 (1 + 1 )(ei 2 − ei1 ) (13) R3 R3 R2
e1 − e2 = (1 +
ei1 + I e1 R1 ei1 I R2 ei2 I ei2 + e2 R1 R4 R3 R3 + e0 R4

Fig. 10 Three op. amp. Instrumentation Amplifier.

Thus by varying R2 very large gain can be achieved, but the relationship is inverse. Since three op. amps. are responsible for achieving this gain, the bandwidth does not suffer. There are many commercially available single chip instrumentation amplifiers in the market. Their gains can be adjusted by connecting an external resistance, or by selecting the gains (50, 100 or 500) through jumper connections.

5.

Concluding Remarks

Several issues have to be taken into consideration for the design of a signal conditioning circuit. Linearity, sensitivity, loading effect, bandwidth, common mode rejection are the important issues that affect the performance of the signal conditioning circuits. In this lesson, we have learnt about different configurations of unbalanced D.C. and A.C bridges, those are suitable for resistive, capacitive and inductive type transducers. Besides the characteristics of different types of amplifiers using common operational amplifiers have also been discussed in details. However, the actual design is dependent on the particular sensing element to be used and its characteristics. Version 2 EE IIT, Kharagpur 13

Several other types of signal conditioning circuits (e.g. phase sensitive detector, filters and many others) have been left out in the discussion.

Problems
1. A resistance temperature detector using copper as the detecting element has a resistance of 100Ω at 0oC. The resistance temperature co-efficient of copper is 0.00427/oC at 0oC. The sensing element is put in an unbalanced wheatstone bridge as in fig.2, the other arms are fixed resistances of 100Ω each. Plot the unbalanced voltage vs. temperature for temperature variation from 0oC to 100oC, if the excitation voltage is E = 2V. Are the characteristics linear or nonlinear? Justify your answer. 2. Explain the advantage of using push-pull arrangement in a bridge circuit. 3. For what arm ratio the sensitivity of an unbalanced wheatstone bridge is maximum? 4. A noninverting amplifier provides higher input impedance to the measuring circuit compared to an inverting amplifier- justify. 5. Define CMRR of an op. amp. Why is it important for designing a measurement system? 6. Design a differential amplifier of gain 10. 7. Discuss the main features of an instrumentation amplifier. 8. A differential amplifier circuit shown in fig. 8 has the resistances: R1 = 10K, R2 = 100K, R3 = 11K and R4 = 100K. Assuming the op. amp. To be an ideal one, find the CMRR of the amplifier. 9. A simple capacitance amplifier circuit is shown in fig. P1. C1 represents a capacitive sensor whose nominal value is 50 pF. C2 is a fixed capacitor of 25 pf. Find the output voltage if the sinusoidal excitation voltage 1V peak-to peak at frequency 1kHz. Assume the op.amp. to be an ideal one.

C1 ei e0 +

Fig. P1.

Answers
1. For 100oC change in temperature is change in resistance for the RTD is 42.7Ω. So the condition ΔR R << 1 is not satisfied. As a result the bridge output is highly nonlinear. 6. Refer fig.8. Any combination of resistances satisfying eqn.(9) and R2 R1 = 10 will do. Typical values, R1 = 10 K and R2 = 100 K . 8. 40.83dB 9. 2 Version 2 EE IIT, Kharagpur 14

Version 2 EE IIT, Kharagpur 15

Module 2
Measurement Systems
Version 2 EE IIT, Kharagpur 1

Lesson 10
Errors and Calibration
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
At the end of this lesson, the student should be able to: • • • • • • Define error Classify different types of errors Define the terms: mean, variance and standard deviation Define the term limiting error for an instrument Estimate the least square straight line from a set of dispersed data Distinguish between the terms: single point calibration and two point calibration.

Introduction
Through measurement, we try to obtain the value of an unknown parameter. However this measured value cannot be the actual or true value. If the measured value is very close to the true value, we call it to be a very accurate measuring system. But before using the measured data for further use, one must have some idea how accurate is the measured data. So error analysis is an integral part of measurement. We should also have clear idea what are the sources of error, how they can be reduced by properly designing the measurement methodology and also by repetitive measurements. These issues have been dwelt upon in this lesson. Besides, for maintaining the accuracy the readings of the measuring instrument are frequently to be compared and adjusted with the reading of another standard instrument. This process is known as calibration. We will also discuss about calibration in details.

Error Analysis
The term error in a measurement is defined as: Error = Instrument reading – true reading. Error is often expressed in percentage as: (1)

% Error =

Instrument reading − true reading X 100 true reading

(2)

The errors in instrument readings may be classified in to three categories as: 1. Gross errors 2. Systematic errors 3. Random Errors. Gross errors arise due to human mistakes, such as, reading of the instrument value before it reaches steady state, mistake of recording the measured data in calculating a derived measured, etc. Parallax error in reading on an analog scale is also is also a source of gross error. Careful reading and recording of the data can reduce the gross errors to a great extent. Systematic errors are those that affect all the readings in a particular fashion. Zero error, and bias of an instrument are examples of systematic errors. On the other hand, there are few errors, Version 2 EE IIT, Kharagpur 3

the cause of which is not clearly known, and they affect the readings in a random way. This type of errors is known as Random error. There is an important difference between the systematic errors and random errors. In most of the case, the systematic errors can be corrected by calibration, whereas the random errors can never be corrected, the can only be reduced by averaging, or error limits can be estimated.

Systematic Errors
Systematic errors may arise due to different reasons. It may be due to the shortcomings of the instrument or the sensor. An instrument may have a zero error, or its output may be varying in a nonlinear fashion with the input, thus deviating from the ideal linear input/output relationship. The amplifier inside the instrument may have input offset voltage and current which will contribute to zero error. Different nonlinearities in the amplifier circuit will also cause error due to nonlinearity. Besides, the systematic error can also be due to improper design of the measuring scheme. It may arise due to the loading effect, improper selection of the sensor or the filter cut off frequency. Systematic errors can be due to environmental effect also. The sensor characteristics may change with temperature or other environmental conditions. The major feature of systematic errors is that the sources of errors are recognisable and can be reduced to a great extent by carefully designing the measuring system and selecting its components. By placing the instrument in a controlled environment may also help in reduction of systematic errors. They can be further reduced by proper and regular calibration of the instrument.

Random Errors
It has been already mentioned that the causes of random errors are not exactly known, so they cannot be eliminated. They can only be reduced and the error ranges can be estimated by using some statistical operations. If we measure the same input variable a number of times, keeping all other factors affecting the measurement same, the same measured value would not be repeated, the consecutive reading would rather differ in a random way. But fortunately, the deviations of the readings normally follow a particular distribution (mostly normal distribution) and we may be able to reduce the error by taking a number of readings and averaging them out. Few terms are often used to chararacterize the distribution of the measurement, namely,

1 n (3) ∑ xi n i =1 where n is the total number of readings and xi is the value of the individual readings. It can be shown that the mean value is the most probable value of a set of readings, and that is why it has a very important role in statistical error analysis. The deviation of the individual readings from the mean value can be obtained as : Mean Value x =
_

Deviation d i = xi − x (4) We now want to have an idea about the deviation, i.e., whether the individual readings are far away from the mean value or not. Unfortunately, the mean of deviation will not serve the purpose, since, _ _ 1 n 1 _ Mean of deviation = ∑ ( xi − x) = x − (n x) = 0 n i =1 n

_

Version 2 EE IIT, Kharagpur 4

So instead, variance or the mean square deviation is used as a measure of the deviation of the set of readings. It is defined as: _ 1 n Variance V = ( xi − x) 2 = σ 2 (5) ∑ n − 1 i =1 The term σ is denoted as standard deviation. It is to be noted that in the above expression, the averaging is done over n-1 readings, instead of n readings. The above definition can be justified, if one considers the fact that if it is averaged over n, the variance would become zero when n=1 and this may lead to some misinterpretation of the observed readings. On the other hand the above definition is more consistent, since the variance is undefined if the number of reading is one. However, for a large number of readings (n>30), one can safely approximate the variance as, _ 1 n Variance V = ∑ ( xi − x) 2 = σ 2 (6) n i =1 The term standard deviation is often used as a measure of uncertainty in a set of measurements. Standard deviation is also used as a measure of quality of an instrument. It has been discussed in Lesson-3 that precision, a measure of reproducibility is expressed in terms of standard deviation.

Propagation of Error
Quite often, a variable is estimated from the measurement of two parameters. A typical example may be the estimation of power of a d.c circuit from the measurement of voltage and current in the circuit. The question is that how to estimate the uncertainty in the estimated variable, if the uncertainties in the measured parameters are known. The problem can be stated mathematically as, y = f ( x1 , x2 ,....., xn ) Let (7) If the uncertainty (or deviation) in xi is known and is equal to Δxi , (i = 1,2,..n) , what is the overall uncertainty in the term y? Differentiating the above expression, and applying Taylor series expansion, we obtain, ∂f ∂f ∂f Δy = Δx1 + Δx2 + ...... + Δxn (8) ∂x1 ∂x2 ∂xn Since Δxi can be either +ve or –ve in sign, the maximum possible error is when all the errors are positive and occurring simultaneously. The term absolute error is defined as, ∂f ∂f ∂f Absolute error : Δy = Δx1 + Δx2 + ...... + Δxn (9) ∂x1 ∂x2 ∂xn But this is a very unlikely phenomenon. In practice, x1 , x2 ,....., xn are independent and all errors do not occur simultaneously. As a result, the above error estimation is very conservative. To alleviate this problem, the cumulative error in y is defined in terms of the standard deviation. Squaring equation (8), we obtain,

⎛ ∂f ⎞ ⎛ ∂f ⎞ ∂f ∂f 2 2 (Δy ) = ⎜ (10) ⎟ ⎜ ⎟ ⎜ ∂x ⎟ (Δx1 ) + ⎜ ∂x ⎟ (Δx2 ) + ...... + 2 ∂x ∂x .(Δx1Δx2 ) + ....... 1 2 ⎝ 2⎠ ⎝ 1⎠ If the variations of x1 , x2 ,..... are independent, positive value of one increment is equally likely to be associated with the negative value of another increment, so that the some of all the cross
2

2

2

Version 2 EE IIT, Kharagpur 5

product terms can be taken as zero, in repeated observations. We have already defined variance V as the mean squared error. So, the mean of (Δy ) 2 for a set of repeated observations, becomes the variance of y, or

⎛ ∂f ⎞ ⎛ ∂f ⎞ ⎟ V ( x1 ) + ⎜ V ( y) = ⎜ ⎟ ⎜ ∂x ⎟ V ( x2 ) + ...... ⎜ ∂x ⎟ ⎝ 2⎠ ⎝ 1⎠ So the standard deviation of the variable y can be expressed as:
2 ⎡⎛ ∂f ⎞ 2 ⎤ ⎛ ∂f ⎞ 2 2 ⎟ σ ( x2 ) + ......⎥ σ ( y ) = ⎢⎜ ⎟ σ ( x1 ) + ⎜ ⎜ ⎟ ⎜ ∂x ⎟ ⎢⎝ ∂x1 ⎠ ⎥ ⎝ 2⎠ ⎣ ⎦ 1 2

2

2

(11)

(12)

Limiting Error
Limiting error is an important parameter used for specifying the accuracy of an instrument. The limiting error (or guarantee error) is specified by the manufacturer to define the maximum limit of the error that may occur in the instrument. Suppose the accuracy of a 0-100V voltmeter is specified as 2% of the full scale range. This implies that the error is guaranteed to be within ± 2V for any reading. If the voltmeter reads 50V, then also the error is also within ± 2V. As a 2 result, the accuracy for this reading will be ×100 = 4% . If the overall performance of a 5 measuring system is dependent on the accuracy of several independent parameters, then the limiting or guarantee error is decided by the absolute error as given in the expression in (9). For example, if we are measuring the value of an unknown resistance element using a wheatstone bridge whose known resistors have specified accuracies of 1%, 2% and 3% respectively, then,

Since Rx =

R1 R2 , we have, R3 R2 R RR ΔR1 + 1 ΔR2 − 1 2 2 ΔR3 R3 R3 R3

ΔRx =
or,

ΔRx ΔR1 ΔR2 ΔR3 = + − Rx R1 R2 R3

Then following the logic given to establish (9), the absolute error is computed by taking the positive values only and the errors will add up; as a result the limiting error for the unknown resistor will be 6%.

Importance of the Arithmetic Mean
It has been a common practice to take a number of measurements and take the arithmetic mean to estimate the average value. But the question may be raised: why mean? The answer is: The most probable value of a set of dispersed data is the arithmetic mean. The statement can be substantiated from the following proof. Let x1 , x2 , x3 ,...., xn be a set of n observed data. Let X be the central value (not yet specified). So the deviations from the central value are ( x1 − X ), ( x2 − X ),....(xn − X ).
Version 2 EE IIT, Kharagpur 6

The sum of the square of the deviations is: S sq = ( x1 − X ) 2 + ( x2 − X ) 2 + ... + ( xn − X ) 2
2 2 = x12 + x2 + .... + xn − 2 X ( x1 + x2 + ... + xn ) + nX 2 So the problem is to find X so that Ssq is minimum. So,

dS sq = −2( x1 + x2 + ... + xn ) + 2nX = 0 dX

or,
X =
_ 1 ( x1 + x2 + ... + xn ) = x n

So the arithmetic mean is the central value in the least square sense. If we take another set of readings, we shall reach at a different mean value. But if we take a large number of readings, definitely we shall come very close to the actual value (or universal mean). So the question is, how to determine the deviations of the different set of mean values obtained from the actual value?

Standard deviation of the mean
Here we shall try to find out the standard deviation of the mean value obtained from the universal mean or actual value. Consider a set of n number of readings, x1 , x2 , x3, ...., xn . The mean value of this set expressed as:
1 ( x1 + x2 + ... + xn ) = f ( x1 + x2 + ... + xn ) n Using (11) for the above expression, we can write: x=
_

⎛ ∂f ⎞ ⎛ ∂f ⎞ ⎛ ∂f ⎞ V ( x) = ⎜ ⎜ ∂x ⎟ V ( x1 ) + ⎜ ∂x ⎟ V ( x2 ) + ...... + ⎜ ∂x ⎟ V ( xn ) ⎟ ⎜ ⎟ ⎜ ⎟ ⎝ 1⎠ ⎝ 2⎠ ⎝ n⎠ 1 = 2 [V ( x1 ) + V ( x2 )..... + V ( xn )] n
_

2

2

2

Now the standard deviation for the readings x1 , x2 ,..., xn is defined as:
⎡1 ⎤ 2 σ = ⎢ [V ( x1 ) + V ( x2 )..... + V ( xn )]⎥ , where n is large. ⎣n ⎦ Therefore, _ 1 σ2 V ( x ) = 2 ( n.σ 2 ) = n n Hence, the standard deviation of the mean,
1

σ ( x) =

_

σ

n
_

(13)

which indicates that the precision can be increased, (i.e. σ ( x) reduced) by taking more number of

observations. But the improvement is slow due to the

n factor.
Version 2 EE IIT, Kharagpur 7

Example: Suppose, a measuring instrument produces a random error whose standard deviation is 1%. How many measurements should be taken and averaged, in order to reduce the standard deviation of the mean to <0.1%? Solution: In this case, 1 σ < 0. 1 ; ∴ n > = 10 ; or , n > 100. 0 .1 n

Least square Curve Fitting
Often while performing experiments, we obtain a set of data relating the input and output variables (e.g. resistance vs. temperature characteristics of a resistive element) and we want to fit a smooth curve joining different experimental points. Mathematically, we want to fit a polynomial over the experimental data, such that the sum of the square of the deviations between the experimental points and the corresponding points of the polynomial is minimum. The technique is known as least square curve fitting. We shall explain the method for a straight line curve fitting. A typical case of least square straight line fitting for a set of dispersed data is shown in Fig. 1. We want to obtain the best fit straight line out of the dispersed data shown.
y

x x x x x x x

0

x

Fig. 1 Least square straight line fitting.

Suppose, we have a set of n observed data ( x1 , y1 ), ( x2 , y2 ),..., ( xn , yn ) . We want to estimate a straight line

y∗ = a0 + a1 x (14) such that the integral square error is minimum. The unknowns in the estimated straight line are the constants a0 and a1 . Now the error in the estimation corresponding to the i-th reading: ei = yi − y ∗ = yi − a0 − a1 xi The integral square error is given by,
Se = ∑ ei2
i =1 n

= ∑ ( yi − a0 − a1 xi ) 2
i =1

n

For minimum integral square error,

Version 2 EE IIT, Kharagpur 8

∂Se ∂Se =0, =0 ∂a0 ∂a1
or,
n ∂Se = − 2∑ ( yi − a0 − a1 xi ) = 0 ∂a0 i =1

(15)

and

n ∂Se = − 2∑ xi ( yi − a0 − a1 xi ) = 0 ∂a1 i =1 From (15) and (16), we obtain,

(16)

∑y
i =1
n i =1 i

n

i

− a 0 .n − a 1 ∑ x i = 0 ,
i =1
n n 0 i =1 i 1 i =1 2 i

n

∑x y −a ∑x −a ∑x
i

=0.

Solving, we obtain,
a1 = or,
_ _ 1 n xi yi − x . y ∑ n a1 = i =1n _2 1 xi2 − x ∑ n i =1

n∑ xi yi − ∑ xi ∑ yi
i =1 i =1 i =1

n

n

n

n∑ xi − (∑ xi ) 2
2 i =1 i =1

n

n

(17)

where x and y are the mean values of the experimental readings xi and yi respectively. Using (14), we can have,
a0 = y − a1 x
_ _

_

_

(18)

Calibration and error reduction
It has already been mentioned that the random errors cannot be eliminated. But by taking a number of readings under the same condition and taking the mean, we can considerably reduce the random errors. In fact, if the number of readings is very large, we can say that the mean value will approach the true value, and thus the error can be made almost zero. For finite number of readings, by using the statistical method of analysis, we can also estimate the range of the measurement error. On the other hand, the systematic errors are well defined, the source of error can be identified easily and once identified, it is possible to eliminate the systematic error. But even for a simple instrument, the systematic errors arise due to a number of causes and it is a tedious process to identify and eliminate all the sources of errors. An attractive alternative is to calibrate the instrument for different known inputs.

Version 2 EE IIT, Kharagpur 9

Calibration is a process where a known input signal or a series of input signals are applied to the measuring system. By comparing the actual input value with the output indication of the system, the overall effect of the systematic errors can be observed. The errors at those calibrating points are then made zero by trimming few adjustable components, by using calibration charts or by using software corrections. Strictly speaking, calibration involves comparing the measured value with the standard instruments derived from comparison with the primary standards kept at Standard Laboratories. In an actual calibrating system for a pressure sensor (say), we not only require a standard pressure measuring device, but also a test-bench, where the desired pressure can be generated at different values. The calibration process of an acceleration measuring device is more difficult, since, the desired acceleration should be generated on a body, the measuring device has to be mounted on it and the actual value of the generated acceleration is measured in some indirect way. The calibration can be done for all the points, and then for actual measurement, the true value can be obtained from a look-up table prepared and stored before hand. This type of calibration, is often referred as software calibration. Alternatively, a more popular way is to calibrate the instrument at one, two or three points of measurement and trim the instrument through independent adjustments, so that, the error at those points would be zero. It is then expected that error for the whole range of measurement would remain within a small range. These types of calibration are known as single-point, two-point and three-point calibration. Typical input-output characteristics of a measuring device under these three calibrations are shown in fig.2. The single-point calibration is often referred as offset adjustment, where the output of the system is forced to be zero under zero input condition. For electronic instruments, often it is done automatically and is the process is known as auto-zero calibration. For most of the field instruments calibration is done at two points, one at zero input and the other at full scale input. Two independent adjustments, normally provided, are known as zero and span adjustments. One important point needs to be mentioned at this juncture. The characteristics of an instrument change with time. So even it is calibrated once, the output may deviate from the calibrated points with time, temperature and other environmental conditions. So the calibration process has to be repeated at regular intervals if one wants that it should give accurate value of the measurand through out.

Output variable

Output variable

Output variable

Ideal

Ideal

Ideal

Actual Input variable (a)

Actual Input variable (b)

Actual Input variable (c)

Fig. 2 (a) single point calibration, (b) two point calibration, (c) three point calibration

Version 2 EE IIT, Kharagpur 10

Conclusion
Errors and calibration are two major issues in measurement. In fact, knowledge on measurement remains incomplete without any comprehensive idea on these two issues. In this chapter we have tried to give a brief overview about errors and calibration. The terms error and limiting error have been defined and explained. The different types of error are also classified. The methods for reducing random errors through repetitive measurements are explained. We have also discussed the least square straight line fitting technique. The propagation of error is also discussed. However, though the importance of mean and standard deviation has been elaborated, for the sake of brevity, the normal distribution, that random errors normally follows, has been left out. The performance of an instrument changes with time and many other physical parameters. In order to ensure that the instrument reading will follow the actual value within reason accuracy, calibration is required at frequent intervals. In this process we compare and adjust the instrument readings to give true values at few selected readings. Different methods of calibration, e.g., single point calibration, two point calibration and three point calibration have been explained.

References
1. 2. 3. 4. M.B.Stout: Basic Electrical Measurements, 2/e, Prentice Hall of India, New Delhi, 1981. R.Pallas-Areny and J.G.Webster: Analog Signal Processing, John Wiley, NY, 1999. R.B. Northrup: Introduction to Instrumentation and Measurements (2/e), CRC Press, Boca Raton, 2005. J.W. Dally, W.F. Riley and K.G. McConnell: Instrumentation for Engineering Measurements (2/e), John Wiley & Sons, NY, 2003.

Review Questions
1. 2. 3. 4. 5. Define error. A temperature indicator reads 189.80C when the actual temperature is 195.50C. Find the percentage error in the reading. Distinguish between gross error and systematic error. Write down two possible sources of systematic error. Explain the term limiting error. In a multiple range instrument it is always advisable to take a reading where the indication is near the full scale: justify. The most probable value of a set of dispersed data is the arithmetic mean: justify. The resistance value at a temperature t of a metal wire, Rt is given by the expression, R t =R 0 (1+αt) where, R 0 is the resistance at 0oC, and α is the resistance temperature coefficient. The resistance values of the metal wire at different temperatures have been tabulated as given below. Obtain the values of R 0 and α using least square straight line fitting. Temperature (oC) Resistance (ohm) 20 107.5 40 117.0 60 117.0 80 128.0 100 142.5

Version 2 EE IIT, Kharagpur 11

6. 7.

Most of the instruments have zero and span adjustments. What type of calibration is it? Explain three point calibration and its advantage over the other types of calibration.

Version 2 EE IIT, Kharagpur 12

Module 3
Process Control
Version 2 EE IIT, Kharagpur 1

Lesson 11
Introduction to Process Control
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
At the end of this lesson, the student should be able to • • • • • • • • • Distinguish with examples the difference between sequential control and continuous process control. Identify three special features of a process. Differentiate between manipulating variable and disturbance. Distinguish between a SISO system and MIMO system and give at least one example in each case. Develop linearised mathematical models of simple systems. Give an example of a time delay system. Identify the parameters on which the time delay is dependent. Sketch the step response of a first order system with time delay. State and explain the significance of transfer function matrix.

1.

Introduction

We often come across the term process indicating a set up or a plant that we want to control. Thus by a process we may mean a unit of chemical plant (say, a distillation column), or a manufacturing system (say, an assembly shop), or a food processing industry and so on. We may want to automate the process; we may also like to control certain parameters of the system output (say, level of a tank, pressure of steam etc.). Broadly speaking, there could be two types of control; we might want to carry out. The first one is called sequential control, where the control action is carried out in a sequence. A good example for this type of operation could be in an automated car manufacturing system, where the assembly of parts is carried out in a sequence (on a conveyor line). Here the control action is sequential in nature and works in a preprogrammed open loop fashion (implying that there is no feedback of the output signal to the controller). Programmable Logic Controller (PLC) is often used to carry out these operations. But there are cases, where the control action needed is continuous in nature and precise control of the output variable is required. Take for example, the drum level control of a boiler. Here, the water level of the drum has to be maintained within a small band, in spite of variations of steam flow rate, steam pressure etc. This type of control is sometimes called modulating control, as the control variable is modulated to keep the process variable at a constant value. Feedback principle is used for these types of control. Now onwards, we would concentrate on the control of these types of processes. But in order to design a controller effectively, we must have a thorough knowledge about the dynamics of the process. A mathematical model of the process dynamics often helps us to understand the process behaviour under different operational conditions. In this lesson, we would discuss the basic characteristics of this type of processes where continuous control is used for controlling certain variables at the outputs.

Version 2 EE IIT, Kharagpur 3

2.

Characteristics of a Process

Different processes have different characteristics. But, broadly speaking, there are certain characteristics features those are more or less common to most of the processes. They are: (i) (ii) (iii) The mathematical model of the process is nonlinear in nature. The process model contains the disturbance input The process model contains the time delay term.

In general a process may have several input variables and several output variables. But only one or two (at most few) of the input variables are used to control the process. These inputs, used for manipulating the process are called manipulating variables. The other inputs those are left uncontrolled are called disturbances. Few outputs are measured and fed back for comparison with the desired set values. The controller operates based on the error values and gives the command for controlling the manipulating variables. The block diagram of such a closed loop process can be drawn as shown in Fig. 1.

In order to understand the behavour of a process, let us take up a simple open loop process as shown in Fig. 2. It is a tank containing certain liquid with an inflow line fitted with a valve V1 and an outflow line fitted with another valve V2. We want to maintain the level of the liquid in the tank; so the measured output variable is the liquid level h. It is evident from Fig.2 that there are two variables, which affect the measured output (henceforth we will call it only output) - the liquid level. These are the throttling of the valves V1 and V2. The valve V1 is in the inlet line, and it is used to vary the inflow rate, depending on the level of the tank. So we can call the inflow rate as the manipulating variable. The outflow rate (or the throttling of the valve V2 ) also affect the level of the tank, but that is decided by the demand, so not in our hand. We call it a disturbance (or sometimes as load). The major feature of this process is that it has a single input (manipulating variable) and a single output (liquid level). So we call it a Single-Input-Single-Output (SISO) process. We would see afterwards that there are Multiple-Input-Multiple-Output (MIMO) processes also.

Version 2 EE IIT, Kharagpur 4

3.

Mathematical Modeling

In order to understand the behaviour of a process, a mathematical description of the dynamic behaviour of the process has to be developed. But unfortunately, the mathematical model of most of the physical processes is nonlinear in nature. On the other hand, most of the tools for analysis, simulation and design of the controllers, assumes, the process dynamics is linear in nature. In order to bridge this gap, the linearization of the nonlinear model is often needed. This linearization is with respect to a particular operating point of the system. In this section we will illustrate the nonlinear mathematical behaviour of a process and the linearization of the model. We will take up the specific example of a simple process described in Fig.2. Let Qi and Qo are the inflow rate and outflow rate (in m3/sec) of the tank, and H is the height of the liquid level at any time instant. We assume that the cross sectional area of the tank be A. In a steady state, both Qi and Qo are same, and the height H of the tank will be constant. But when they are unequal, we can write, dH (1) Qi − Qo = A dt But the outflow rate Qo is dependent on the height of the tank. Considering the Valve V2 as an orifice, we can write, (please refer eqn.(4) in Lesson 7 for details)

Qo =

Cd A2 1− β 4

2g

γ

( P − P2 ) 1

(2)

We can also assume that the outlet pressure P2=0 (atmospheric pressure) and P = ρ gH 1

(3)

Considering that the opening of the orifice (valve V2 position) remains same throughout the operation, equation (2) can be simplified as: Qo = C H (4) Where, C is a constant. So from equation (1) we can write that, dH (5) Qi − C H = A dt The nonlinear nature of the process dynamics is evident from eqn.(5), due to the presence of the term H . Version 2 EE IIT, Kharagpur 5

In order to linearise the model and obtain a transfer function between the input and output, let us assume that initially Qi =Qo =Qs; and the liquid level has attained a steady state value Hs. Now suppose the inflow rate has slightly changed, then how the height will change? Now expanding Qo in Taylor’s series, we can have:

Qo = Qo ( H s ) + Qo ( H s ) ( H − H s ) + ..... Taking the first order approximation, from eqn.(4), Qo ( H s ) = C H s = Qs Qo ( H s ) =

(6)

C 2 Hs Then from (1) and (6), we can write, d (H − H s ) C dH =A (7) Qi − Qs − (H − H s ) = A dt dt 2 Hs Now, we define the variables q and h, as the deviations from the steady state values, q = Qi − Qs (8) h = H − Hs We can write from (7), dh 1 (9) q=A + h dt R 2 Hs (10) Where, R = C It can be easily seen, that eqn.(9) is a linear differential equation. So the transfer function of the process can easily be obtained as: h( s ) R (11) = q(s) τ s + 1 Where, τ = RA .

It is to be noted that all the input and output variables in the transfer function model represent, the deviations from the steady state values. If the operating point (the steady state level Hs in the present case) changes, the parameters of the process (R and τ ) will also change. The importance of linearisation needs to be emphasized at this juncture. The mathematical models of most of the physical processes are nonlinear in nature; but most of the tools for design and analysis are for linear systems only. As a result, it is easier to design and evaluate the performance of a system if its mathematical model is available in linear form. Linearised model is an approximation of the actual model of the system, but it is preferred in order to have a physical insight of the system behaviour. It is to be kept in mind that this model is valid as long as the variation of the variables around the operating point is small. There are few systems whose dynamic behaviour is highly nonlinear and it is almost impossible to have a linear model of a system. For example, it is possible to develop the linearised transfer function model of an a.c. servomotor, but it is not possible for a step motor. Referring to Fig. 2, if the valve V1 is motorized and operated by electrical signal, we can also develop the model relating the electrical input signal and the output. Again, we have so far assumed that the opening of the valve V2 to be constant, during the operation. But if we also
Version 2 EE IIT, Kharagpur 6

consider its variation, that would also affect the dynamics of the tank model. So, the effect of disturbance can be incorporated in the overall plant model, as shown in Fig.3, by introducing a disturbance transfer function D(s). D(s) can be easily by using the same methodology as described earlier in this section.

4.

Higher Order System Model

We have considered a single tank and developed the linearised model of it. So it has a single time constant τ . But there are more complex processes. If there are two tanks coupled together, as shown in Fig.4, then we would have two time constants τ 1 and τ 2 . But it is evident that the dynamics of two tanks are coupled. Considering the change in the inflow q(t) as the input and the change in the level of the second tank h2 (t ) as the output variable, with a little bit of calculation, it can be shown that the transfer function of this coupled tank system is, h2 ( s ) R2 = (12) 2 q( s) τ 1τ 1s + (τ 1 + τ 2 + A1R2 ) s + 1 The constants are similar to the earlier section with added suffixes corresponding to tank 1 and tank 2 respectively. In this case we have neglected the effects of the disturbances.

5.

Time delay

It has been mentioned earlier that one of the major characteristics of a process is the presence of time delay. This time delay term is often referred as “transportation lag”, since it is generated due to the delay in transportation of the output to the measuring point. The presence and effect of time delay can be easily explained with an example of a simple heat exchanger, as shown in Fig. 5.

Version 2 EE IIT, Kharagpur 7

In this case the transfer of heat takes place between the steam in the jacket and water in the tank. The measured output is the water temperature at the outlet T(t). For controlling this temperature, we may vary the steam flow rate at its inlet. So the manipulating variable is the steam flow rate. We can also identify a number of input variables those act as the disturbance, thus affecting the temperature at the water outlet; for example, inlet steam temperature, inlet water temperature and the water flow rate. The temperature transducer should be placed at a location in the water outlet line just after the tank (location A in Fig. 5). But suppose, due to the space constraint, the transducer was placed at location B, at a distance L from the tank. In that case, there would be a delay sensing this temperature. If T(t) is the temperature measured at location A, then the temperature measured at location B would be T (t − τ d ) . The time delay term τ d can be expressed in terms of the physical parameters as: τd = L v (13) where L is the distance of the pipeline between locations A and B; and v is the velocity of water through the pipeline. Noting from the Laplace Transformation table,

L f (t − τ d ) = e − sτ F ( s )
d

(14)

we can conclude that an additional term of e − sτ d would be introduced in the transfer function of the system due to the time delay factor. Thus the transfer function of an ordinary first order plant with time delay is Ke − sτ d G (s) = 1 + sτ and its step response to a unit step input is as shown in Fig. 6.

Version 2 EE IIT, Kharagpur 8

It can be seen that though the input has been at t = 0, the output remains zero till t = τ d . This time delay present in the system may often be the main cause for instability of a closed loop system operation.

6.

Multiple Input Multiple Output Systems

So far we have considered the behviour of single input single output (SISO) systems only. In these cases, we had a single manipulating variable to control a single output variable. But in many cases, we have a number of inputs to control a number of outputs simultaneously, and the input-outputs are not decoupled. This will be evident if we consider a system, slightly modified system from that one shown in Fig. 4. In the modified system, we have added another inlet flow line in tank 2, as shown in Fig. 7.

If we consider the changes in inflow rates q1 and q2 are in inputs and the changes in the liquid levels of the two tanks h1 and h2 as the outputs, then the complete input-output behaviour can be modeled using the transfer function matrix, as shown below:
⎡ h1 ( s ) ⎤ ⎡G11 ( s ) ⎢ h ( s ) ⎥ = ⎢G ( s ) ⎣ 2 ⎦ ⎣ 21 G12 ( s ) ⎤ ⎡ q1 ( s ) ⎤ G22 ( s ) ⎥ ⎢ q2 ( s ) ⎥ ⎦ ⎦⎣

(15)

We define G ( s ) as the transfer function matrix and G12 ( s ) ⎤ ⎡G11 ( s ) G ( s) = ⎢ G22 ( s ) ⎥ ⎣G21 ( s ) ⎦

Version 2 EE IIT, Kharagpur 9

In general, if there are m inputs and p outputs, then the order of the transfer function matrix is p X m. The MIMO system can also be further classified depending on the number of inputs and outputs. If the number of inputs is more than the number of outputs (m>p), then the system is called an overactuated system. If the number of inputs is less than the number of outputs (m<p), then the system is an underactuated system; while they are equal then the system is square (implying the G(s) is a square matrix). A Multi-input-multi-output nonlinear system can be described in its state variable form as:
x = f ( x, u ) y = g ( x, u ) where x is the state vector, u is the input vector and y is the output vector. f and g are nonlinear functions of x and u. The above nonlinear system can also be linearised over its operating point and can be described in the state-space form as:

x = Ax + Bu (16) y = Cx + Du where u is the input vector of dimension m; y is the output vector of dimension p and x is an ndimensional vector representing the states. The transfer function matrix G ( s ) can be obtained as: −1 (17) G ( s ) = C ( sI − A ) B + D

For more details, please refer any book on Control Systems.

Example -1
A typical example of a MIMO process and development of its model has been taken up in this example. This is related to the control of temperature and humidity in order to maintain an artificial tropical climate inside a room in winter. Fresh cold air is forced inside the room through a fan. Steam is added to the air in order to maintain the humidity. An electric heater is used to increase the temperature of the room. So the control inputs are the steam valve position θ (t ) and the voltage applied to the heater v(t). The measured outputs are room temperature c(t) and the relative humidity h(t). The detail process description is shown in Fig. 8.

Version 2 EE IIT, Kharagpur 10

The room humidity is controlled by manipulating the valve position in steam input. The dynamics of the humidity inside the room can be expressed as:

τθ

dh + h(t ) = K hθ (t ) dt

(18)

On the other hand, both the steam flow rate and the voltage to the heater contribute in deciding the temperature of the room. The dynamics of the room temperature can be expressed by the equations as:

dcθ + cθ (t ) = Kθ θ (t ) dt dc τ v v + cv (t ) = K v v(t ) dt c(t ) = cθ (t ) + cv (t )

τθ

(19)

Taking the Laplace transformation, the overall system dynamics can be written in form of the transfer function matrix as:

⎡ kh ⎡ h( s) ⎤ ⎢ sτ θ + 1 ⎢ ⎢c( s ) ⎥ = ⎢ k ⎣ ⎦ θ ⎢ sτ + 1 ⎣ θ

⎤ ⎥ ⎡θ ( s) ⎤ ⎥ kv ⎥ ⎢ v ( s ) ⎥ ⎦ ⎣ ⎥ sτ v + 1 ⎦ 0

(20)

7.

Conclusion

In this lesson, a brief introduction about several aspects of process control has been provided. By process control, we mean continuous control of one or few parameters of the process output, and feedback control is used for maintaining these values. This type of operation is distinctly different from “sequential control” that is basically discrete in operation, and open loop in nature. In order to effectively control the process, a thorough knowledge about the characteristics of the process is needed. We have learnt about few terms like, disturbance, time delay etc., whose behaviours affect the performance and stability to a great extent. Again, the mathematical model of the process is often nonlinear in nature, but most of the design and analysis tools are suitable for linear systems only. As a result the system model needs to be linearised over an operating point. The basic techniques for developing the mathematical model and its linarisation are elaborated in this lesson. Many of the process control systems are Single-Input-Single-Output (SISO) type. But there are cases where, a number of outputs are to be simultaneously controlled by manipulating a number of inputs. This requirement leads to Multiple-Input-Multiple-Output (MIMO) systems. Typical examples of MIMO processes and their mathematical models have been discussed in this lesson. So far, we have refrained from discussing about the controllers. Different types of controllers, their performances and tuning of the controller parameters would be taken up in the subsequent lessons.

Version 2 EE IIT, Kharagpur 11

References
1. 2. 3. 4. D.R. Coughanowr: Process systems analysis and control (2/e), McgrawHill, NY, 1991. B. Liptak; Process Control: Instrument Engineers Handbook K. Ogata: Modern Control engineering (2/e), Prentice Hall of India, new Delhi, 1995. G. Stephanopoulos: Chemical Process Control, Prentice Hall of India, New Delhi, 1996.

Review Questions
1. Take the example of a simple liquid level control system for a vessel. Draw the block diagram for the closed loop control. Identify, the input, output, manipulating variable and disturbance for this case. Explain the physical reason behind generation of time delay. Why time delay is not so prevalent in electrical systems? Justify. The transfer function of a process is given by: 5 e −0.5 s ; where time is expressed in minutes. Sketch the open loop response for a G (s) = 1+ s unit step input to the process. What are the factors the time delay is dependent on? Give an example of a multiple-input-multiple-output process. What do you mean by transfer function matrix? What is its relation with the state space description of a system? The dynamic equation of a system is given by: dx 3 + x2 = 4 dt Is it a linear system, or a nonlinear system? Justify. If it is a nonlinear system, then linearise the equation for a steady state operating condition xs=2.

2. 3.

4. 5. 6. 7.

Version 2 EE IIT, Kharagpur 12

Module 3
Process Control
Version 2 EE IIT, Kharagpur 1

Lesson 12
P-I-D Control
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
At the end of this lesson, the student should be able to: • • • • • • • • • Write the input-output relationship of a P-I-D controller Explain the improvement of transient response in closed loop with P-controller Explain the presence of offset in presence of simple P-controller Define Proportional Band Explain the elimination of steady state error with Integral Control. Define the error transfer function and compute steady state error Explain the advantages of P-I controller over simple P and I actions Explain the effect of P-D controller Recommend a suitable controller configuration for a particular process.

Introduction
In the last lesson, a brief introduction about a process control system has been given. The basic control loop can be simplified for a single-input-single-output (SISO) system as in Fig.1. Here we are neglecting any disturbance present in the system.

The controller may have different structures. Different design methodologies are there for designing the controller in order to achieve desired performance level. But the most popular among them is Proportional-Integral-derivative (PID) type controller. In fact more than 95% of the industrial controllers are of PID type. As is evident from its name, the output of the PID controller u(t) can be expressed in terms of the input e(t), as:
t ⎡ ⎤ de(t ) 1 + ∫ e(τ ) dτ ⎥ u (t ) = K p ⎢ e ( t ) + τ d dt τi 0 ⎣ ⎦

(1)

and the transfer function of the controller is given by:

⎛ 1 ⎞ C ( s) = K p ⎜1 + τ d s + ⎟ τis ⎠ ⎝
The terms of the controller are defined as: K p = Proportional gain

(2)

τ d = Derivative time, and
Version 2 EE IIT, Kharagpur 3

τ i = Integral time.
In the following sections we shall try to understand the effects of the individual componentsproportional, derivative and integral on the closed loop response of this system. For the sake of simplicity, we consider the transfer function of the plant as a simple first order system without time delay as:

P(s) =

K 1+τ s

(3)

Proportional control
With the proportional control action only, the closed loop system looks like:

Now the closed loop transfer function can be expressed as:

KK p KK p KK p c( s ) 1 = 1+τ s = = KK p 1 + KK p + τ s 1 + KK p 1 + τ ' s r (s) 1+ 1+τ s τ '
1 + KK p

(4)

where τ =

.

For a step input r ( s ) =

A , s

c( s) =

A 1 + KK p s 1 + τ ' s

KK p

(

)
(5)

or,

AKK p ⎛ − sτ ' ⎞ τ c(t ) = ⎟ ⎜1 − e 1 + KK p ⎝ ⎠

The system response is shown in Fig. 2.

Version 2 EE IIT, Kharagpur

4

From eqn. (5) and Fig. 2, it is apparent that: 1. 2.
1 (i.e. the time constant decreases). 1 + KK p There is a steady state offset between the desired response and the output response = ⎛ KK p ⎞ A A ⎜1 − = . ⎜ 1 + KK p ⎟ 1 + KK p ⎟ ⎝ ⎠

The time response improves by a factor

This offset can be reduced by increasing the proportional gain; but that may also cause increase oscillations for higher order systems. The offset, often termed as “steady state error” can also be obtained from the error transfer function and the error function e(t) can be expressed in terms of the Laplace transformation form: 1 A 1+τ s A e( s ) = = KK p s 1 + KK p + τ s s 1+ 1+τ s Using the final value theorem, the steady state error is given by:

ess = Lt e ( t ) = Lt s e( s) = Lt
t →∞ s →0

1+τ s A A = s →0 1 + KK + τ s s 1 + KK p p

Often, the proportional gain term, Kp is expressed in terms of “Proportional Band”. It is inversely proportional to the gain and expressed in percentage. For example, if the gain is 2, the proportional band is 50%. Strictly speaking, proportional band is defined as the %error to move the control valve from fully closed to fully opened condition. However, the meaning of this statement would be clear to the reader afterwards.

Integral Control
If we consider the integral action of the controller only, the closed loop system for the same process is represented by the block diagram as shown in Fig. 3.

Version 2 EE IIT, Kharagpur

5

Proceeding in the same way as in eqn. (4), in this case, we obtain,

K τ s(1 + τ s) c( s ) K = i = K r (s) 1 + K + τ i s + ττ i s 2 τ i s(1 + τ s)
From the first observation, it can be seen that with integral controller, the order of the closed loop system increases by one. This increase in order may cause instability of the closed loop system, if the process is of higher order dynamics. A For a step input r ( s ) = , s τ i s (1 + τ s ) A A 1 e( s ) = = K s τ i s (1 + τ s ) + K s 1+ τ i (1 + τ s )

ess = Lt s e( s ) = 0
s →0

So the major advantage of this integral control action is that the steady state error due to step input reduces to zero. But simultaneously, the system response is generally slow, oscillatory and unless properly designed, sometimes even unstable. The step response of this closed loop system with integral action is shown in Fig. 4.

Proportional Plus Integral (P-I) Control
With P-I controller the block diagram of the closed loop system with the same process is given in Fig. 5.

Version 2 EE IIT, Kharagpur

6

It is evident from the above discussions that the P-I action provides the dual advantages of fast response due to P-action and the zero steady state error due to I-action. The error transfer function of the above system can be expressed as:

e( s ) = r ( s)

τ i s(1 + τ s) 1 = 2 KK p (1 + τ i s) s ττ i + (1 + KK p )τ i s + KK p 1+ τ i s(1 + τ s)

In the same way as in integral control, we can conclude that the steady state error would be zero for P-I action. Besides, the closed loop characteristics equation for P-I action is:

s 2ττ i + (1 + KK p )τ i s + KK p = 0 ;
from which we can obtain, the damping constant as: ⎛ 1 + KK p ⎞ τi ξ =⎜ ⎟ 2 ⎠ KK pτ ⎝ whereas, for simple integral control the damping constant is: ⎛1⎞ τ ξ =⎜ ⎟ i ⎝ 2 ⎠ Kτ Comparing these two, one can easily observe that, by varying the term Kp, the damping constant can be increased. So we can conclude that by using P-I control, the steady state error can be brought down to zero, and simultaneously, the transient response can be improved. The output responses due to (i) P, (ii) I and (iii) P-I control for the same plant can be compared from the sketch shown in Fig. 6.

Proportional Plus Derivative (P-D) Control
The transfer function of a P-D controller is given by: C ( s ) = K p (1 + τ d s ) Version 2 EE IIT, Kharagpur 7

P-D control for the process transfer function P ( s ) =

K apparently is not very useful, since it 1+τ s

cannot reduce the steady state error to zero. But for higher order processes, it can be shown that the stability of the closed loop system can be improved using P-D controller. For this, let us take 1 up the process transfer function as P ( s ) = 2 . Looking at Fig.7, we can easily conclude that Js with proportional control, the closed loop transfer function is Kp 2 Kp c( s) = Js = 2 K r (s) Js + K p 1+ p 2 Js and the characteristics equation is Js 2 + K p = 0 ; giving oscillatory response. But with P-D controller, the closed loop transfer function is: K p (1 + τ d s ) K p (1 + τ d s ) c( s) Js 2 = = 2 K (1 + τ s ) Js + K p (1 + τ d s ) r (s) 1+ p 2 d Js whose characteristics equation is Js 2 + K pτ d s + K p = 0 ; that will give a stable closed loop response.

The step responses of this process with P and P-D controllers are compared in Fig.8.

Version 2 EE IIT, Kharagpur

8

Proportional-Integral-Derivative (PID) control
It is clear from above discussions that a suitable combination of proportional, integral and derivative actions can provide all the desired performances of a closed loop system. The transfer function of a P-I-D controller is given by:

⎛ 1 ⎞ C ( s) = K p ⎜1 + τ d s + ⎟ τis ⎠ ⎝
The order of the controller is low, but this controller has universal applicability; it can be used in any type of SISO system, e.g. linear, nonlinear, time delay etc. Many of the MIMO systems are first decoupled into several SISO loops and PID controllers are designed for each loop. PID controllers have also been found to be robust, and that is the reason, it finds wide acceptability for industrial processes. However, for proper use, a controller has to be tuned for a particular process; i.e. selection of P,I,D parameters are very important and process dependent. Unless the parameters are properly chosen, a controller may cause instability to the closed loop system. The method of tuning of P,I,D parameters would be taken up in the next lesson. It is not always necessary that all the features of proportional, derivative and integral actions should be incorporated in the controller. In fact, in most of the cases, a simple P-I structure will suffice. A general guideline for selection of Controller mode, as suggested by Liptak [1], is given below.

Guideline for selection of controller mode
1. Proportional Controller: It is simple regulating type; tuning is easy. But it normally introduces steady state error. It is recommended for process transfer functions having a pole at origin, or for transfer functions having a single dominating pole; for example with

P( s ) =

K ; with τ 1 (1 + sτ 1 )(1 + sτ 2 )(1 + sτ 3 )

τ 2 ,τ 1

τ3 .

2. Integral Control: It does not exhibit steady state error, but is relatively slow responding. It is particularly effective for: (i) very fast process, with high noise level (ii) process dominated by dead time (iii) high order system with all time constants of the same magnitude. 3. Proportional plus Integral (P-I) Control: It does not cause offset associated with proportional control. It also yields much faster response than integral action alone. It is widely used for process industries for controlling variables like level, flow, pressure, etc., those do not have large time constants. 4. Proportional plus Derivative (P-D) Control: It is effective for systems having large number of time constants. It results in a more rapid response and less offset than is possible by pure proportional control. But one must be careful while using derivative action in control of very fast processes, or if the measurement is noisy (e.g. flow measurement). 5. Proportional plus Integral plus Derivative (P-I-D) Control: It finds universal application. But proper tuning of the controller is difficult. It is particularly useful for controlling slow variables, like pH, temperature, etc. in process industries. Version 2 EE IIT, Kharagpur 9

Conclusion
In this lesson, the basic functions of a P-I-D controller have been explained. Most of the industrial controllers are P-I-D in nature. The major reasons behind the popularity of P-I-D controller are its simplicity in structure and the appilicability to variety of processes. Moreover the controller can be tuned for a process, even without detailed mathematical model of the process. However, proper tuning of the controller parameters requires extensive experimentation. The methods for controller tuning would be discussed in the next lesson. Crudely speaking, the desired closed loop performances, such as fast response, zero steady state error and less overshoot are achieved through incorporation of P,I and D actions respectively. But the choice of P-D, P-I or P-I-D structure depends on the type of the process we intend to control. A brief guideline for selection of controller is provided in this lesson. There are few more issues those need to be addressed while using P-I controller. The most important among them is the anti-windup control. Further details about anti-windup would be discussed in Lesson 14.

References
1. B. Liptak: Process Control: Instrument Engineers Handbook

Review Questions
1. A P-I controller has a proportional band of 50% and integration time of 2sec. Find the transfer function of the controller. 2. The transfer function of a first order plant is G ( s ) =
2 . It is used in a unity feedback 1 + 2s system as shown in Fig. 2 with a proportional controller of proportional band 100%. Find the steady state error for a unit step input, and the time constant of the closed loop system.

3. Repeat problem 2 if proportional band is 50%. 4. What would be the steady state error for the plant in problem-2 if the transfer function of 1 the controller is Gc ( s) = 2(1 + ) ? 2s 5. Incorporation of P-I action may lead to instability in the closed loop performance- justify. 6. How does incorporation of derivative action in the controller improve the closed loop performance? 7. Why derivative control is not recommended for a flow control process? 8. What type of controller would you recommend for control of pH level in a liquid?

Version 2 EE IIT, Kharagpur 10

Module 3
Process Control
Version 2 EE IIT, Kharagpur 1

Lesson 13
Controller Tuning
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
At the end of this lesson, the student should be able to • • • • Explain the importance of tuning of controller for a particular process Name the three experimental techniques for controller tuning Explain the three methods for tuning of P, I and D parameters Explain the terms: Auto Tuning, Bumpless Transfer and Integration Wind Up.

1.

Introduction

The importance of P-I-D controller and the features of P, I and D actions were elaborated in the last lesson. It was also mentioned that the controller could be easily incorporated in a process, whatever be the type of a process: linear or nonlinear, having dead time or not. It is needless to say that the controller parameters influence heavily the performance of the closed loop system. Again, the choice of the value of the P, I and D parameters is very much process dependent. As a result, thorough knowledge about the plant dynamics is important for selection of these parameters. In most of the cases, it is difficult to obtain the exact mathematical model of the plant. So, we have to rely on the experimentation for finding out the optimum settings of the controller for a particular process. The process of experimentation for obtaining the optimum values of the controller parameters with respect to a particular process is known as controller tuning. It is needless to say, that controller tuning is very much process dependent and any improper selection of the controller settings may lead to instability, or deterioration of the performance of the closed loop system. In 1942 two practicing engineers, J.G. Ziegler and N.B. Nichols, after carrying out extensive experiments with different types of processes proposed certain tuning rules, there were readily accepted and till now are used as basic guidelines for tuning of PID controllers. Subsequently, G.H. Cohen and G.A. Coon in 1953 proposed further modifications of the above techniques. Still then, the methods are commonly known as Ziegler-Nichols method. Substantial amount of research has been carried out on tuning of P-I-D controllers since last six decades. Several other methods have also been proposed. Most of them are model based, i.e. they assume that the mathematical model of the system is available to the designer. In fact, if the mathematical model of the system is available, many of them perform better than conventional Ziegler-Nichols method. But the strength of the ZN method is that it does not require a mathematical model, but controller parameters can simply be chosen by experimentation. We would be discussing the three experimental techniques those come under the commonly known Ziegler-Nichols method. Now let us look back to whatever discussed in lessons 11 and 12. The closed loop system can be described as shown in Fig. 1.

Version 2 EE IIT, Kharagpur 3

The error signal is fed to the controller and the controller provides output u(t). Since the capacity of the controller to deliver output power is limited, an actuator is needed in between the controller and the process, which will actuate the control signal. It may be a valve positioner to open or close a valve; or a damper positioner to control the airflow through a damper. The controller considered here is a P-I-D controller whose input and output relationship is given by the equation:
t ⎡ ⎤ de(t ) 1 + ∫ e(τ ) dτ ⎥ u (t ) = K p ⎢ e(t ) + τ d τi 0 dt ⎣ ⎦

Our objective is to find out the optimum settings of the P,I,D parameters, namely K p , τ d and τ i through experimentation, which will provide satisfactory closed loop performance, of the particular process in terms of, say, stability, overshoot, setting time etc. Three methods of tuning are elaborated in the following sections.

2.

Reaction Curve Technique

This is basically an open loop technique of tuning. Here the process is assumed to be a stable first order system with time delay. The closed loop system is broken as shown in Fig.2; a step input is applies at m ' , output is measured at b. In fact, a bias input may be necessary so that the plant output initially becomes close to the nominal value. The step input is superimposed on this bias value. The input and the output response are plotted by suitable means as shown in Fig. 3.

Version 2 EE IIT, Kharagpur 4

M,L and K are measured. Let us define the following terms corresponding to Fig. 2: Slope = N, Time Constant T=K/N Lag Ratio R=L/T Then, the recommended optimum settings, for P, P-I and P-I-D controller are as follows.

Optimum settings
P-Control:

Kp =

M R (1 + ) 3 NL

P-I Control:

Kp =

M 9 R ( + ); NL 10 12
M 4 R ( + ); NL 3 4

⎛ 30 + 3R ⎞ τ i = L⎜ ⎟ ⎝ 9 + 20 R ⎠
⎛ 32 + 6R ⎞ τ i = L⎜ ⎟ ⎝ 13 + 8R ⎠

P-I-D Control: K p =

⎛ 4 ⎞ τ d = L⎜ ⎟ ⎝ 11 + 2R ⎠

3.

Closed Loop Technique (Continuous Cycling method)

The major objection to the tuning methodology using reaction curve technique is that process has to be run in open loop that may not always be permissible. For tuning the controller when the process is in under closed loop operation, there are two methodologies. The first one, continuous cycling method is explained below.

Version 2 EE IIT, Kharagpur 5

Referring Fig.1, the loop is closed with the controller output connected to the actuator input. Here, the controller is first set to P-mode, making τ d = 0 and τ i = ∞ . The proportional gain Kp is increased gradually to

K p = K p max ,

till the system just starts oscillating with constant

amplitude continuously. The output waveform is plotted as shown in Fig.4. The time period of continuous oscillation Tu is noted. The recommended optimum settings are: P Control:

K p = 0.5K p max

P-I Control: K p = 0.45K p max , τi = P-I-D Control: K p = 0.6 K p max ,

Tu 1.2

τi =

Tu T , τd = u 2 8

Points to Ponder
a) b) Why is the proportional gain Kp for PI control is less than the value for P-only control? Why Kp for PID control is more than that PI?

4.

Closed Loop Technique (Damped oscillation method)

In many cases, plants are not allowed to undergo through sustained oscillations, as is the case for tuning using continuous cycling method. Damped oscillation method is preferred for these cases. Here, initially the closed loop system is operated initially with low gain proportional control mode with τ d = 0 and τ i = ∞ . The gain is increased slowly till a decay ratio (p2/p1) of 1/4th is obtained in the step response in the output, as shown in Fig. 5. Under this condition, the period of damped oscillation, Td is also noted. Let Kd be the proportional gain setting for obtaining 1/4th decay ratio. The optimum settings for a P-I-D controller are:

K p = Kd ; τ i =

Td T ;τd = d 6 1.5

Version 2 EE IIT, Kharagpur 6

5.

General comments about controller tuning

The different methodologies of controller tuning, known as Ziegler-Nichols method have been illustrated in the earlier sections. It is to be remembered that the recommended settings are empirical in nature, and obtained from extensive experimentation with number of different processes; there is no theoretical basis behind these selections. As a result, a better combination of the P, I, D values may always be found, that will give less oscillation and better settling time. But with no a-priori knowledge of the system, it is always advisable to perform the experimentation and select the controller settings, obtained from Ziegler-Nichols method. But there is always scope for improving the performance of the controller by fine-tuning. So, ZieglerNichols method provides initial settings that will give satisfactory, result, but it is always advisable to fine-tune the controller further for the particular process and better performance is expected to be achieved. Nowadays digital computers are replacing the conventional analog controllers. P-I-D control actions are generated through digital computations. Digital outputs of the controllers are converted to analog signals before they are fed to the actuators. In many cases, commercial software are available for Auto tuning the process. Here the controller generates several commands those are fed to the plant. After observing the output responses, the controller parameters are selected, similar to the cases discussed above.

6.

Integration windup and Bumpless transfer

Two major issues of concern with the close loop operation with P-I-D controllers are the Integration Windup and the requirement of providing Bumpless Transfer. These two issues are briefly elaborated below. The methodologies for providing Anti-integration Windup and Bumpless Transfer would be discussed in the next lesson.

Integration Windup
A significant problem with integral action is that when the error signal is large for a significant period of time. This can occur every time when there is large change in set point. If there is a sudden large change in set point, the error will be large and the integrator output in a P-I-D control will build up with time. As a result, the controller output may exceed the saturation limit of the actuator. This windup, unless prevented may cause continuous oscillation of the process that is not desirable. Version 2 EE IIT, Kharagpur 7

Bumpless Transfer
When a controller is switched from manual mode to auto-mode, it is desired that the input of the process should not change suddenly. But since there is always a possibility that the decision of the manual mode of control and the auto mode of control be different, there may be a sudden change in the output of the controller, giving rise to a sudden jerk in the process operation. Special precautions are taken for bumpless transfer from manual to auto-mode.

References
1. B. Liptak: Process Control: Instrument Engineers Handbook 2. D.R. Coughanowr: Process systems analysis and control (2/e), McgrawHill, NY, 1991. 3. D. Eckman: Process Control, Wiley, NY, 1958.

Review Questions
1. What does controller tuning mean? 2. Name the three techniques for controller tuning, those are commonly known as ZieglerNichols method. 3. Explain the reaction curve technique for tuning of controller. What are its limitations? 4. What do you mean by Auto Tuning? Explain briefly. 5. What is meant by Bumpless Transfer? 6. Why provision for Anti- integration Windup is necessary for process with P-I-D control?

Answers to Points to Ponder
a) Addition of integral control action to P-only control tends to make the closed loop system more oscillatory; in order to overcome this problem, the suggested value of Kp with ZN tuning is reduced. b) Addition of derivative action again damps down the oscillation; as a result larger value of Kp in a PID controller is permissible.

Version 2 EE IIT, Kharagpur 8

Module 3
Process Control
Version 2 EE IIT, Kharagpur 1

Lesson 14
Implementation of P-I-D Controllers
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
At the end of this lesson, the student should be able to • • • • • • Suggest a method to achieve Bumpless transfer Suggest two methods for prevention of Integration Windup Explain a scheme for implementation of pneumatic P-I controller Explain a scheme for implementation of P-I-D controller using electronic circuit Distinguish between position algorithm and velocity algorithm for implementation of digital P-I-D controller Explain the advantages of using velocity algorithm over position algorithm

Introduction
We have discussed in the last lesson the tuning rules for PID controllers. In this lesson, we shall discuss about how to implement a PID controller in an actual system. Looking back to the history of the PID controller, the PID controllers in the initial days were all pneumatic. In fact, all the experimentation by Ziegler and Nichols were carried out with pneumatic controllers. But pneumatic controllers were slow in nature. After the development of electronic devices and operational amplifiers, the electronic controllers started replacing the conventional pneumatic controllers. But with the advent of the microprocessors and microcontrollers, the focus of development is now towards the implementation with digital PID controllers. The major advantage of using digital PID controllers is that the controllers parameter can be programmed easily; as a result, they can be changed without changing any hardware. Moreover, the same digital computer can be used for a number of other applications besides generating the control action. In this lesson, we shall first discuss about the PID controller implementation with pneumatic and electronic components and then discuss about the different algorithms those can be used for digital implementation of PID controllers. But before that, we shall discuss about the different schemes of implementation of bumpless transfer and antiwindup actions. These two issues were described in the last lesson. We shall now see, how they can be implemented in actual practice.

Bumpless Transfer
It is quite normal to set up some processes using manual control initially, and once the process is close to normal operating point, the control is transferred to automatic mode through auto/manual switch. In such cases, in order to avoid any jerk in the process the controller output immediately after the changeover should be identical to the output set in the manual mode. This can be achieved by forcing the integral output at the instant of transfer to balance the proportional and derivative outputs against the previous manual output; i.e. Integral output = {(previous manual) – (proportional + derivative) output}. Similarly, for automatic to manual transfer, initially the manual output is set equal to the controller output and the difference is gradually reduced by incrementing or decrementing the manual output to the final value of the manual signal and thus effecting a change over. Version 2 EE IIT, Kharagpur 3

Another way to transfer from Auto to Manual mode in a bumpless manner, the set point may be made equal to the present value of the process variable and then slowly changing the set point to its desired value. The above features can be easily be implemented if a digital computer is used as a controller. This provision eliminates the chance of the process receiving sudden jolt during transfer.

Prevention of Integration Windup
The effect of integration windup has been discussed in the last lesson. If there is a sudden large change in set point, the error between the set point and the process output will suddenly shoot up and the integrator output due to this error will build up with time. As a result, the controller output may exceed the saturation limit of the actuator. This windup, unless prevented may cause continuous oscillation of the process. There exits several methods through which integration windup can be prevented. Before we go to the actual methods, let us consider the input-output characteristics of an actuator as shown in Fig. 1. Its characteristics is similar to of an amplifier, where the output varies linearly with the input till the input is within a certain range; beyond that the output becomes constant either at the maximum or the minimum values of the output. The upper and lower limits of the output may correspond to the flow rates of a control valve when the valve is at fully open and fully closed position.

The first method uses a switch to break the integral action, whenever the actuator goes to saturation. This can be illustrated by Fig. 2. Consider schematic arrangement of a controller shown in the figure. When the switch is closed, transfer function of the controller can be obtained as:

Version 2 EE IIT, Kharagpur 4

⎛ ⎞ ⎜ ⎟ ⎛ 1+τis ⎞ u ( s) 1 1 ⎟ = Kp ⎜ = Kp ⎜ ⎟ = K p (1 + ) 1 ⎟ e( s ) τis ⎜ 1− ⎝ τis ⎠ ⎜ τ s +1 ⎟ i ⎝ ⎠
So when the switch is closed, the controller acts as a P-I controller. On the other hand, if the switch is open, it is a simple P- controller. The switch is activated by the position of the actuator. If the actuator is operating in the linear range, the switch is closed, and the controller is in P-I mode. But whenever the actuator is in the saturation mode, the switch is automatically opened; the controller becomes a P-controller. As a result, any windup due to the presence of integral mode is avoided. Another technique for antiwindup action is illustrated in Fig.3. Here we assume that the slope of the actuator in the linear range is unity. As a result, when the actuator is operating in the linear range the error eA is zero, and the controller acts as a PI controller. But when the actuator is in saturation mode, the error eA is negative for a positive e. This will reduce the integral action in the overall control loop.

Other anti-windup precautions are: (i) (ii) Closing the I-action only when the error is small (say 5% to 10% of the range), Limiting the output of the I-action block.

However, application of these techniques require an intimate knowledge of the plant behaviour.

Pneumatic Controller
It has been already mentioned that the early days PID controllers were all pneumatic type. The advantage of pneumatic controllers is its ruggedness, while its major limitation is its slow response. Besides it requires clean and constant pressure air supply. The major components of a pneumatic controller are bellows, flapper nozzle amplifier, air relay and restrictors (valves). The integral and derivative actions are generated by controlling the passage of air flow through restrictors to the bellows. However, the details of the scheme for generation of PID action in a pneumatic controller will be elaborated in Lesson 29 and 30. A simple scheme for

Version 2 EE IIT, Kharagpur 5

implementation of a pneumatic PI controller is shown in Fig. 4. Details explanation will be provided in Lessons 29 and 30.

Here four bellows are connected to a force beam as shown. The measured process variable is converted to air pressure and connected to the bellows P1. Similarly the air pressure corresponding to the set point signal is applied to the bellow P2. The error corresponding to the measured value and the set point generates a force on the left hand side of the force beam. There is an adjustable pivot arrangement that sets the proportional gain of the amplifier. The right hand side of the force beam is connected to two bellows, P3 and P4 and a flapper nozzle amplifier. The output air pressure is dependent on the gap between the flapper and nozzle. An air relay enhances the air handling capacity. The output pressure is directly fed back to the feedback bellows P4, and also to P3 through a restrictor (valve). The opening of this restrictor decides the integral action to be applied. With a slight modification of this scheme, a pneumatic PID controller can also be implemented.

Electronic PID Controllers
Electronic PID controllers can be obtained using operational amplifiers and passive components like resistors and capacitors. A typical scheme is shown in Fig. 5. With little calculations, it can be shown that the circuit is capable of delivering the PID actions as:

⎡ 1 de(t ) ⎤ e0 (t ) = K p ⎢e(t ) + ∫ e(τ ) dτ + τ d ⎥ τi dt ⎦ ⎣

(1)

Version 2 EE IIT, Kharagpur 6

R2 of the first amplifier; R1 the integral action is decided by R3 and C1 and the derivative action by R5 and C2. The final output however comes out with a negative sign, compared to eqn. (1) (though the positive sign can also be obtained by using a noninverting amplifier at the input stage, instead of the inverting amplifier). The op. amps. Shown in the circuits are assumed to be ideal.
It is evident from Fig. 5, the proportional gain Kp is decided by the ratio

Digital P-I-D Control
In the digital control mode, the error signal is first sampled and the controller output is computed numerically through a digital processor. Now Controller output for a continuous-type P-I-D controller:

⎡ 1 de( t ) ⎤ u ( t ) = K p ⎢e( t ) + ∫ e(τ) dτ + τ d ⎥ dt ⎦ τi ⎣

(1)

The above equation can be discretised at small sampling interval T0 as shown in Fig. 6.

Version 2 EE IIT, Kharagpur 7

Taking the first order derivative,

de dt

1 [e(k ) − e(k − 1)] T0
k −1 i =0

and using rectangular integration, we can approximate as:

∫ e(τ)dt ⇒ T0 ∑ e(i) ; t = kT0
0

t

Now replacing the derivative and integral terms in eqn. (1), one can obtain,

⎡ T u (k ) = K p ⎢e(k ) + 0 τi ⎣

⎤ τ ∑ e(i) + T {e(k) − e(k − 1)}⎥
k −1 i =0 d 0

(2)

The above algorithm is known as Position algorithm. But the major problem here is that the error values at all the time instants are to be stored (or at least the second term of the r.h.s of Eqn. (2) at each instant have to be stored). An alternative approach known as velocity algorithm can be obtained as follows. From (2), one can write the error signal at the (k-1) th instant as:

⎡ ⎤ τ T0 k −2 u (k − 1) = K p ⎢e(k − 1) + ∑ e(i) + d {e(k − 1) − e(k − 2)}⎥ τ i i =0 T0 ⎣ ⎦
Subtracting eqn. (3) from (2), we can have:

(3)

Δu (k ) = u (k ) − u (k − 1) ⎡ ⎤ T τ = K p ⎢e(k ) − e(k − 1) + 0 e(k − 1) + d {e(k ) − 2e(k − 1) + e(k − 2)}⎥ τi T0 ⎣ ⎦
= q 0 e(k ) + q1e(k − 1) + q 2 e(k − 2)
where, (4)

Version 2 EE IIT, Kharagpur 8

q 0 = K p (1 +

τd
T0

)

q1 = − K p (1 + q2 = K p

2τ d T0 − ) T0 τi

τd
T0

The above algorithm is known as Velocity algorithm. The major advantage of this algorithm is that it is of recursive type. It calculates the incremental output at each sample instant. As a result, it requires only to store three previous values: e(k), e(k-1) and e(k-2). Besides it has got several other advantages also those are elaborated below:

1.

Bumpless Transfer

During the transfer from manual to auto mode, it is desired that the input command to the process should not change suddenly. In Position algorithm, due to the difference between the set point and the output variable, it is always possible that the existing error will wind up and the value of ∑ e(k ) be large when the switching from manual to auto mode takes place. This will cause a large change in the input u(k) in the auto mode. But in the velocity algorithm, this will be prevented, since it provides only incremental change in input [u(k)-u(k-1)]. This will lead to bumpless transfer.

2.

Prevention of Integration Windup

If there is a sudden change in set point, the error will increase continuously to take the value of ∑ e(k ) in position algorithm to a large value. Afterwards, even if the error reduces to zero, or changes sign,

∑ e(k ) will

take a large time to come to zero, or change sign, resulting in

integration windup. But in velocity algorithm, as soon as the error changes sign, term T corresponding to the integration 0 e(k − 1) (in eqn. (4)) will change sign. Thus even if the

τI

actuator is saturated, it will come back to linear range within one sampling period.

3.

Protection against Computer Failure

Another advantage of velocity algorithm is its ability to protect the process in case of computer failure. In case there is a failure of the computer, there will be no increment or decrement of the control input, and it will retain the last value before the computer failure, thus preventing process failure. However, there are certain pitfalls of velocity algorithm also. In case of presence of noise in the measurement of the error value at a particular sampling instant, the controller will immediately act, taking it to be a signal. But in position algorithm, the integration term

∑ e(i) will prevent
i =0

k −1

Version 2 EE IIT, Kharagpur 9

such a quick action. Some times, a digital filter with low pass characteristics is used to filter out the unwanted noise before it reaches the controller input.

Conclusion
Actual implementation of PID controllers is not a trivial task. In this lesson several methods for implementation of the control actions, as well as prevention of certain undesirable effects while the controller is in use in the process have been discussed. The implementation of the PID action can be carried out using pneumatic or electronic discrete components (hydraulic controllers are also in use, but to a limited extent). The basic schematics of pneumatic and electronic PID controllers have been explained. Next, the implementation of the control action using a digital controller has been discussed. Two algorithms for digital implementation have been explained. Their relative merits and demerits are also been elaborated. The necessity for using schemes for bumpless transfer and anti-integration windup been explained and few such schematic arrangements have been presented.

References
1. D.P. Eckman: Automatic process Control, Wiley Eastern, New Delhi, 1958. 2. G. Stephanopoulus: Chemical process Control, Prentice Hall of India, New Delhi, 1995. 3. C. Johnson: Process Control Instrumentation Technology (4/e), Prentice Hall of India, New Delhi, 1996. 4. J.M. Jacob: Industrial Control Electronics, Prentice Hall International, NJ, 1989.

Review Questions
1. What problem is envisaged when a controller is switched from manual control mode to auto mode? Suggest a scheme for overcoming this problem. 2. Explain how the saturation of the actuator affects the performance of a PID controller. Suggest one scheme for overcoming the problem. Explain its operation. 3. Draw the schematic diagram of an electronic PI controller. How the proportional and integral constants can be adjusted. 4. What are the major drawbacks of a pneumatic controller? 5. Explain how a continuous time PID control law can be discretised using velocity algorithm. 6. What are the advantages of velocity algorithm, compared to the position algorithm for digital implementation of PID controllers? 7. Given a continuous time PID controller with Kp=3, τ i = 5 min, and τ d = 1 min. Determine the parameters of the corresponding discrete difference equation using velocity algorithm. Assume sampling time = 30 secs.

Answer
1. q0 = 9, q1 = −14.7, q2 = 6 in eqn. (4). Version 2 EE IIT, Kharagpur 10

Module 3
Process Control
Version 2 EE IIT, Kharagpur 1

Lesson 15
Special Control Structures: Feedforward and Ratio Control
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
At the end of this lesson, the student should be able to • • • • • Justify the use of feedforward controller in addition to conventional feedback controller. Draw the block diagram of a feedforward-feedback controller. Find the transfer function of the feedforward controller for complete disturbance rejection. Write down three typical applications of ratio control Give two possible arrangements for achieving ratio control

Introduction
In the last few lessons we have discussed about the different aspects of PID controllers, their tuning, implementation etc. In all the cases the control action considered was feedback type; i.e. the output was fed back and compared with the set point and the error was fed to the controller. The output from the controller was used to control the manipulating variable. However there are several cases, where apart from the feedback action few other control structures are incorporated in order to satisfy certain requirements. In this lesson we would take up two such special control structures: feedforward and ratio control.

Feedforward Control
When the disturbance is measurable, feedforward control is an effective means for cancelling the effects of disturbance on the system output. This is advantageous, since in a simple feedback system, the corrective action starts after the effect of disturbance is reflected at the output. On the other hand, in feedforward control the change in disturbance signal is measured and the corrective action takes place immediately. As a result, the speed and performance of the overall system improves, if feedforward control, together with feedback action is employed. In order to illustrate the effect of feedforward control, let us consider the heat exchange process shown in Fig.1. The cold water comes from a tank and flows to the heat exchanger. The flow rate of cold water can be considered as a disturbance. The change in input flow line may occur due to the change in water level in the tank. Suppose, the feedforward line is not connected, and the controller acts as a feedback control only. If the water inlet flow rate increases, the temperature of the outlet hot water flow will decrease. This will be sensed by the temperature sensor that will compare with the set point temperature and the temperature controller will send signal to open the control valve to allow more steam at the steam inlet. The whole operation is a time consuming and as a result the response of the controller due to the disturbance (inlet water flow rate) is normally slow. But if we measure the change in inlet flow rate by a flowmeter and feed this information to the controller, the controller can immediately take the correcting action anticipating the change in outlet temperature. This will improve the speed of response. Thus feedforward action, in addition to the feedback control improves the performance of the system, but provided, the disturbance is measurable.

Version 2 EE IIT, Kharagpur 3

Let us now draw the block diagram of the overall control operation of the system shown in Fig. 1. The block diagram representation is shown in Fig. 2.

In general, the structure of the feedforward-feedback action in terms of the block diagram of transfer functions can be represented as shown in Fig. 3. Where, G ( s ) = Transfer function of the process (manipulating variable to output) Gn ( s) = Disturbance transfer function (disturbance to output)

Gc ( s) = Transfer function of feedback controller
Version 2 EE IIT, Kharagpur 4

G f ( s) = Transfer function of the feedforward controller
So there are two controllers, one is the conventional feedback controller Gc ( s) , while the other is the feedforward controller that is intended to nullify the effect of disturbance at the output. From Fig. 3, the overall output is:

C(s) = G c ( s ) G(s) E(s) + G n ( s ) N(s)

= G c ( s ) G(s) [ R(s) - C(s) +G f N(s) ] + G n (s) N(s) =G c ( s ) G(s) [ R(s) - C(s) ] + [ G c ( s ) G(s) G f (s) +G n (s) ] N(s)

If we want to select the feedforard transfer function, such that, the effect of disturbance at the output is zero, then we require the co-efficient of N(s) in above equation to be set to zero. Thus, G c ( s ) G(s) G f (s) +G n (s) = 0 G n (s) (1) or, G f (s) = G c (s) G(s) It is to be noted that complete disturbance rejection can be obtained if the transfer functions Gp(s) and Gn(s) are known accurately, which is not possible in actual situation. As a result, the performance of the feedforward controller will deteriorate and complete disturbance rejection can not be achieved, through the effect can be reduced considerably. However the feedback controller would reduce the residual error due to imperfect feedforward control and at the output, the effect of imperfect cancellation may not be felt.

Example -1
Consider the composition control system of a certain reagent. The block diagram of the feedforward-feedback control system is shown in Fig.4. Here we are using a P-I controller for feedback action. Ci is the disturbance signal. Compared to Fig.3, the feedforward controller

Version 2 EE IIT, Kharagpur 5

output in the present case is added after the feedback controller. Suppose, the plant transfer function and the disturbance transfer function are given by: 1 G p (s) = (s+1)3 1 G n (s) = (5s+1) (s+1)3 With a little calculation using eqn.(1), we obtain the transfer function of the feedforward controller as: 1 G f (s) = − 5s+1

But in many cases, practical implementation of the feedforward controller as obtained from eqn.(1) is difficult. It can be easily seen in the above example if the disturbance transfer function is : 1 G n (s) = (s+1) 2 In this case, the feedforward controller transfer function becomes: G f (s) = −(s+1) But since the above transfer function is not a proper one, practical realization of this transfer function is difficult. In these cases, the transfer function of the feedforward controller is chosen in the form of a lag-lead compensator as: K (1+T1s) G f (s) = − f , with T2 <<T1 and Kf is a constant. (1+T2s) In the present case the transfer function G f (s) = −(s+1) can be realized by approximating as: s+1 G f (s) = − . 0.1s+1 In normal case, the transfer function of the process is not known exactly. So the question is how to tune the parameters of the feedforward controller for optimum performance. This can be carried out by performing some experimentation for tuning the controller. The details of the tuning procedure for feedforward controllers can be found in [2].
Version 2 EE IIT, Kharagpur 6

Ratio Control
Ratio control is a special type of feedforward control where the disturbance is measured and the ratio of the process output and the disturbance is held constant. It is mostly used to control the ratio of flow rates of two streams. Flow rates of both the stream are measured, but only one of them is controlled. There can be many examples of application of ratio control. Few examples are: 1. fuel-air ratio control in burners, 2. control of ratio of two reactants entering a reactor at a desired ratio, 3. maintaining the ratio of two blended streams constant in order to maintain the composition of the blend at the desired value. There can be two schemes for achieving ratio control. The first scheme is shown in Fig. 5. In this configuration the ratio of flow rates of two streams is measured and compared with the desired ratio. The error is fed to the controller and the controller output is used to control the flow rate of stream B.

The second possible scheme for ratio control is shown in Fig. 6. Suppose the flow rate of fluid B has to be maintained at a constant fraction of flow rate of fluid A, irrespective of variation of flow rate of A (qA). In this scheme the flow rate of fluid A is multiplied with the desired ratio (set externally) that gives the desired flow rate of fluid B. This is compared with the actual flow rate of fluid B and fed to the controller that operates the control valve.

Version 2 EE IIT, Kharagpur 7

Suppose that the above scheme (Fig. 6) is used for controlling the fuel-air ratio in a burner where airflow rate (fluid B) is controlled. But the desired ratio is also dependent on the temperature of the air. So an auxiliary measurement is needed to measure the temperature of air and set the desired ratio. Such a scheme is shown in Fig. 7.

Version 2 EE IIT, Kharagpur 8

The controllers in ratio control are usually P-I type. This is in order to achieve zero steady state error for maintaining the desired ratio. Derivative action is avoided because the flow is always noisy.

Example -2
We have mentioned earlier that ratio control is a special type of feedforward control. We shall elaborate this point in the following example [4]. Effects of control actions in feedback only mode, feedforward only mode and feedforward-feedback mode will also be elaborated in the context of control of a blending process. Consider a blending system where two streams are blended; one is uncontrolled wild stream that acts as disturbance and another is controlled and acts as the manipulated variable. The ratio of these two streams affect the quality of the output of the process.

Version 2 EE IIT, Kharagpur 9

First consider a simple feedback system shown in Fig. 8(a). The composition of the output stream is unmeasured and fed to the composition controller that controls the flow rate of the manipulated stream. But the performance of this system may not be satisfactory, since any variation of the wild stream will cause the change in composition of the output, that will be felt after certain time. As a result the performance of the feedback controller is sluggish. Version 2 EE IIT, Kharagpur 10

On the other hand the performance of the blending process can be made considerably faster if he incorporate the ratio control scheme as shown in Fig. 8(b). In this case, the disturbance (flow rate of wild stream) is measured and corrective action is taken immediately by controlling manipulated stream so as to maintain a constant ratio between them. It is evident that this the control action is feedforward only, since the composition of the output stream is not unmeasured. But the above scheme also suffers from certain limitations, since the composition variation due to other disturbances is not taken into account. So though faster, if may not yield the desired performance. The performance may be further improved by choosing a feedforward-feedback control scheme (Fig. 8(c)) where the composition of the output stream is measured and the composition controller sets the desired ratio to be maintained.

References
1. 2. 3. 4. G. Stephanopoulus: Chemical process Control, Prentice Hall of India, New Delhi, 1995. D.R. Coughanowr: Process systems analysis and control (2/e), McgrawHill, NY, 1991. K. Ogata: Modern Control Engineering (2/e), Prentice Hall of India, New Delhi, 1995. W.L. Luyben and M.L. Luyben: Essentials of Process Control, McgrawHill, NY, 1997.

Review Questions
1. 2. 3. 4. 5. When would you recommend the use of feedforward controller? What is the advantage of using this control? Draw the general block diagram of a feedforward-feedback control scheme and develop the transfer function of the feedback controller. What modification of the transfer function of the feedforward controller is suggested when the obtained transfer function of the controller is not proper? Explain with an example the principle of ratio control. Elaborate with a block diagram any one scheme for achieving ratio control. Why the controller used for ratio control is normally P-I type?

Version 2 EE IIT, Kharagpur 11

Module 3
Process Control
Version 2 EE IIT, Kharagpur 1

Lesson 16
Special Control Structures: Predictive Control, Control of Systems with Inverse Response
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
At the end of this lesson, the student should be able to • • • • • • Explain with an example the difficulty in controlling a process with dead time. Draw and explain the function of Smith Predictor Compensation Scheme. Explain the two schemes for predictive control in automatic gage control of a rolling mill. Given an example of a process with inverse response. Write down the transfer function of process with inverse response and sketch its step response. Suggest a suitable compensation scheme for control of a process with inverse response.

Predictive Control
It has already been mentioned earlier, how in a chemical process control transportation lag (time delay) comes into picture affecting the model and performance of the process. Model of a − sτ process having a time delay τ d will have a term e d in its numerator of the transfer function. Processes having large time delays are normally difficult to control. A change in set point or disturbance does not reach the output until a time τ d has elapsed. As a result, performance of the closed loop control system is normally sluggish and any change in set point or disturbance will give rise to large oscillations of the output before coming to a steady state value. In order to improve the closed loop performances of such time delay systems, O.J.M. Smith in 1957 first suggested a modification of conventional PID control schemes for processes having large time delay. The scheme for taking a predictive action in presence of transportation delay in the system is better known as Smith Predictor. Let us consider that the transfer function of the process is given by:

G p (s)=e-sτd G(s)

(1)

where G(s) represents the system model without the delay. The basic scheme for Smith Predictor is shown in Fig. 1. Here G(s) is the conventional PID controller designed fro the process G(s). If the system model is exact, the output of the comparator-A would be zero and the outer loop can be ignored. The closed loop system can be simplified as the block diagram shown in Fig. 2. Since the time delay part is absent here, the controller will see the effect of control action much earlier and the sluggish response of the system will improve. The outer loop comes into play if the model of the process is not exact (normally that is expected). Fig. 1 can be redrawn as shown in Fig. 3 with the actual controller is shown inside the dashed line. The performance of the Smith Predictor depends heavily on the actual knowledge of the plant model. Any change in the plant characteristics (particularly dead time) should be instantly taken care of, or otherwise, can cause deterioration of its performance. Besides, the hardware implementation of the controller using analog circuits is difficult. But the control scheme can be implemented with relative ease with a digital controller.

Version 2 EE IIT, Kharagpur 3

Application of Predictive Control in Gage Control of Steel Rolling Mills
In this section we shall give two examples of Predictive Control in control of gage thickness in steel rolling mills. Automatic Gage Control (AGC) is the most important control scheme in a rolling mill. Its main objective is to maintain the thickness of the sheet steel coming out of a rolling mill constant. The basic feedback scheme for AGC in a single stand rolling mill is shown in Fig. 4. The gage is controlled by varying the gap between the rollers in a stand. In fact, there are a number of stands in a rolling milling and the rolling is carried out in stages. Hydraulic actuators are normally used for roll gap adjustment. In this scheme, the gage of the strip is measured at the exit stand and compared with the reference gage command. The error is amplified to operate the servo valve of the hydraulic actuator. The basic control scheme shown here is P-type. Nucleonic detector is used for measurement of gage thickness at the exit stand. Version 2 EE IIT, Kharagpur 4

But because of the fact that the thickness measuring device is installed at a distance from the roll, it will introduce transportation lag in the closed loop system. There will be considerable time lag to detect the variation of sheet thickness at the roll stand and that will lead to oscillatory and unsatisfactory behaviour of closed loop gage control.

In order to improve the performance of the closed loop scheme, Predictive Control mechanism are added with the basic feedback scheme. Here the actual roll gap (and subsequently the gage thickness) is estimated at the roll stand. This can be achieved by two possible methods: • • estimating the roll gap estimating the gap thickness from constant mass flow principle.

In both the cases, the gap is estimated through an approximate model of the rolling stand.

Smith Predictor by estimating the roll gap
In this method, the gage thickness is predicted by estimating the roll gap using the expression:

h = C0 +

P Ks

(2)

where h is the exit thickness of the rolled product, C0 is the no-load roll gap, P is the roll force in the hydraulic actuator and Ks is the structural stiffness of the mill. The overall scheme can be represented as in Fig. 5.

Version 2 EE IIT, Kharagpur 5

Smith Predictor based on Constant Mass Flow principle
The thickness of the strip can also be predicted from constant mass flow principle. Consider a multi-stand mill. Let hi-1 and vi-1 be the thickness and the velocity of the strip at the inlet of the ith stand (i.e. outlet at the (i-1)th stand). So hi and vi are the corresponding parameters at the outlet of the i-th stand. Then from the principle of conservation of mass, we can write, Rate of mass flow in = Rate of mass flow out Considering the width of the strip to be same before and after the stand, we can write:

h i-1 vi-1 = h i vi
or,

hi =

h i-1 vi-1 vi

(3)

The velocities at the inlet and outlet of the stand can be measured by using a Laser Doppler velocity meter, or in a more conventional way by measuring the roll speeds at the two stands. However, some more correction factors are incorporated while predicting the exact gage thickness using constant mass flow principle. A scheme similar to Fig. 5 can be used here in the Smith Predictor scheme.

Systems with Inverse Response
We have seen different types of system response so far: first order system, second order system, system with time lag etc. Another type of system can also be classified with its typical step response pattern: system with inverse response. It is essentially a system whose transfer function is having a zero on the right half plane. This type of system is also called, nonminimum phase system.

Version 2 EE IIT, Kharagpur 6

Example of a system with inverse response
Before going into the details, let us consider a simple example of a system with inverse response. Consider the dynamic characteristics of a boiler drum in a water tube boiler of a steam power plant. High-pressure feedwater is pumped to the drum. Water from the drum circulates through the boiler tubes, gets heated and is converted to steam. This steam again comes back to the drum and subsequently is taken out through the steam flow line to the turbine. So the drum is filled up partially with water and partially with steam, both at high pressure. It is very important to control the water level of the drum at a desired level, by controlling the feedwater flow, with the varying demand of steam. The schematic arrangement can be shown as in Fig. 6.

The instantaneous level of water at the boiler drum is decided by the steam flow rate and the feedwater flow rate and it would reach a steady state when both are equal. Now suppose, the steam flow rate suddenly decreases, the feedwater flow rate remaining constant. At a first glance, it would appear that the drum level should rise. But actually the drum level will initially drop for some time before it rises and reaches a steady value. This is because of the fact that drop in steam flow rate will initially cause the rise in steam pressure in the drum. Due to the rise in pressure, the bubbles present inside the water in the drum will momentarily shrink. This will cause the temporary fall in the drum level. Similarly, for a sudden increase in steam flow rate, the drum level will momentarily swell before it drops down to a steady state vale. A typical response curve of the drum level due to the sudden fall in steam level is plotted in Fig. 7.

Version 2 EE IIT, Kharagpur 7

Transfer function of a system with inverse response
The above response curve can be looked into as a combination of two first order process responses in opposition as shown in Fig. 8. The block diagram representation is shown in Fig. 9.

Version 2 EE IIT, Kharagpur 8

The overall output response can be written as:
⎡ K K ⎤ c(s)= ⎢ 1 - 2 ⎥ u(s) ⎣1+τ1s 1+τ 2s ⎦

or.

c(s)=

(K1τ 2 -K 2 τ1 )s+(K1 -K 2 ) u(s) (1+τ1s)(1+τ 2s)

(4)

Let us assume

τ1 K1 > >1 τ2 K 2

(5)

Then for a step input, the response of process-2 will dominate during the initial phase, but ultimately, process- 1 will dominate in the steady state. It can also be seen that under these two conditions, the system has a zero on the right half of s-plane (it is evident from the fact that K1>K2 and K1 τ 2 <K2 τ1 ). The inverse response is a typical characteristics of a nonminimum phase system, i.e. a system with an unstable zero.

Control of a System with Inverse Response
It is evident that control of a system with inverse response is difficult as in the case of control of a system with large time lag. For a closed loop time lag system with a simple feedback controller, the controller does not see any effect of control action till a time τ d has elapsed. On the other hand, for an inverse system, the controller will see an opposite effect to the expected one. So naturally, some special arrangement, similar to a Smith Predictor scheme is needed for control of inverse systems. Such an arrangement is shown in Fig. 10. This also requires an apriori knowledge of the system model.

From Fig. 10, if we neglect the additional compensator loop with Gc(s) then it is evident that the feedback information received, Version 2 EE IIT, Kharagpur 9

y(s) = C(s)

(K1τ 2 -K 2 τ1 )s+(K1 -K 2 ) e(s) (1+τ1s)(1+τ 2s)

(6)

that has a zero in the right half plane. To eliminate the effect of inverse response, one additional measurement signal must be added that excludes the information of inverse response. That can be achieved by the loop through the compensator Gc(s), that gives an additional output

ys (s)=C(s)G c (s)e(s) ⎛ 1 1 ⎞ = C(s) k ⎜ ⎟ e(s) ⎝ 1+τ 2s 1+τ1s ⎠ Combining (6) and (7), we can have,
y 0 (s) = y ( s ) + ys (s) = C(s) (K1τ 2 -K 2 τ1 )s+k(τ1 -τ 2 )s+(K1 -K 2 ) e(s) (1+τ1s)(1+τ 2s)

(7)

(8)

Now the loop transfer function, i.e. transfer function between yo and e, will have a zero on the left half of s-plane, if the coefficient of s in the numerator of (8) is positive, i.e. using (6)

k>

K 2 τ1 -K1τ 2 τ1 -τ 2

(9)

Thus it is evident that due the presence of the compensator, the inverse response behaviour (i.e. r.h.p zero) will not be felt by the controller. It can be also easily seen that the compensator in Fig. 10 has a similar configuration as in Fig. 6 for smith predictor. In other words, the compensator in Smith Predictor predicts the dead time behaviour of the process, while in the present case; the compensator Gc(s) predicts the inverse behaviour of the process. The basic controller is normally chosen of P-I type. Further improvement of compensator design for an inverse system has been reported in [4].

References
1. G. Stephanopoulus: Chemical process Control, Prentice Hall of India, New Delhi, 1995. 2. K.J. Astrom and B.J. Wittenmark: Computer Controlled Systems, Prentice Hall of India, New Delhi, 1994. 3. V.B. Ginzburg: High Quality Steel Rolling: Theory and practice, Marcel Dekker Inc., NY, 1993. 4. W. Zhang, X. Xu and Y. Sun: Quantitative Performance Design for Inverse-Response Processes, Ind. Eng. Chem. Res. , vol. 39, pp. 2056-2061, 2000.

Review Questions
1. Draw the basic scheme of a Smith Predictor for controlling a process with a transportation lag and explain its principle of operation. 2. For proper operation of a Smith predictor, the model of the process must be known with sufficient accuracy- justify. 3. Explain a Smith predictor scheme used for automatic gage control in rolling mill.
Version 2 EE IIT, Kharagpur 10

4. What do you mean by a system with inverse system? Give an example. 5. A transfer function of a process with inverse response will have a zero on the right half splane – justify the statement. 6. Draw and explain the operation of a compensator for compensating the inverse response of a process in a feedback control scheme.

Version 2 EE IIT, Kharagpur 11

Module 3
Process Control
Version 2 EE IIT, Kharagpur 1

Lesson 17
Special Control Structures: Cascade, Override and Split Range Control
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
At the end of this lesson, the student should be able to • • • • • State two advantage of using cascade control Draw the block diagram representation of cascade control system Write down the governing equations for determining the stability of a cascade control system. Illustrate with an example the use of override control Illustrate with an example the use of split range control.

Introduction
In last two lessons, we have discussed several special control structures those are commonly in use for process control. We shall conclude this part with three more structures. Cascade, override and split range control. Cascade control is widely used for minimizing the effects of certain types of disturbance in the process. On the other hand, override and split range controls are used for controlling certain types of multi variable processes.

Cascade Control
Consider the heat exchange process shown in Fig. 1.Steam is used to heat the water in the heat exchanger. Feedback temperature controller is used to compare the water outlet temperature with the set point and control the steam flow rate by opening or closing the control valve. However due to the change in upstream steam pressure (Pi), the steam flow rate may change, though the control valve is at the same position. This will affect the amount of heat exchanged and the temperature at the water outlet. It will take some time to detect the change in temperature and take subsequent corrective action. In a cascade control, this problem can be overcome by measuring the disturbance (change in flowrate in steam due to upstream pressure variation and a corrective action is taken to maintain constant flowrate of steam. There is an additional controller (flow controller) whose set point is decided by the temperature controller. The schematic arrangement of cascade control is shown in Fig.2. its block diagram is shown in Fig.3. Clearly, there are two control loops – outer and inner, and two controllers. The set point of the inner loop controller is decided by the outer loop primary controller.

Version 2 EE IIT, Kharagpur 3

Version 2 EE IIT, Kharagpur 4

Broadly speaking, there are two major functions of cascade control: (1) to eliminate the effect of some disturbances, and (2) to improve the dynamic performance of the control loop. It is evident from Fig. 3, that the effect of disturbances arising within the inner loop (or secondary loop, as it is called sometimes) is corrected by the secondary controller, before it can influence the output, and the primary controller takes care of the other disturbances in the outer loop. As a result the transient response of the overall system improves. Fig. 4 depicts typical responses of a closed loop system with (a) simple P-I type feedback controller and (b) cascade with primary P-I controller.

There are few other advantages of cascade control. They can be summarized as: 1. A strong (high gain) inner loop reduces the sensitivity and nonlinearity of the plant in the closed loop. The outer loop therefore experiences less parameter perturbations. 2. Cascading makes the use of feedforward control more systematic. In the heat exchanger example (Fig.2) it is possible to measure the water flow and add feedforward compensation to the flow controller. This would improve the speed of response to fluctuation in water flow, which is a disturbance (refer Lesson 15). Version 2 EE IIT, Kharagpur 5

3. Cascade control often makes it possible to use simpler control action than what could be needed for a single controller. Though the number of tunable parameters is more in cascade control, a systematic tuning procedure (inner to outer) is available. On the other hand the major disadvantages of cascade control are (i) more sensors and transmitters and (ii) more number of controllers. For proper operation of the cascade control arrangement, the dynamics of the inner loop (control valve in the present case) should be considerably faster than the dynamics of the outer loop process (heat exchanger in the present case). The inner loop secondary controller is normally chosen as a simple P-type controller with high gain, while the outer loop primary controller is a conventional P-I controller. Due to the presence of two loops in cascade control, the stability of the overall system has to be looked into more carefully. Considering the different elements of the arrangement to be linear, the overall feedback control system can be drawn in terms of the transfer functions as shown in Fig. 5. Here GC1 and GC2 are the two controllers. GP1 is the transfer function of there primary process and GP2 is there transfer function of the valve dynamics.

For maintaining stability of the inner loop the roots of the characteristics equation:

1 + G C2G P2 = 0

(1)

must be on the left hand side of the s-plane. On the other hand, for maintaining the stability of the primary control loop, the roots of the characteristics equation:

⎛ G G ⎞ 1+G C1 ⎜ C2 P2 ⎟ G P1 =0 ⎝ 1+G C2G P2 ⎠

(2)

must be on the left hand side of the s-plane. It is often advantageous to make the inner loop 3 to 4 times faster than the outer loop. The cascade control scheme discussed in this scheme is called a series cascade control. Besides, there is another scheme called parallel cascade control. However, the second one would not been discussed in this lesson.

Version 2 EE IIT, Kharagpur 6

Override Control
In several processes, there may be a single manipulating variable and several output variables. So the control loop should monitor more than just one control variable. Override control (or a selective control, as it is sometimes called), is a special type of multivariable control, where the manipulating variable is controlled by one output variable at a time. Normally only one of the output variables is controlled; but it has also to be ensured that the other output variables do not cross the safe limits. If it is so, a second controller takes over the controller through a selector switch. This can be achieved by using “High Selector Switch” (HSS) or a “Low Selector Switch” (LSS) as required. HSS is used when it is desired that a variable should not exceed an upper limit. Similarly the LSS. Let us consider a simple example of override control. Consider a boiler system shown in Fig. 6. Under normal circumstances, the steam pressure of the boiler is controlled by controlling the flow through the discharge line. The pressure is maintained through the pressure transmitter and the pressure controller. But the water level of the boiler should also not fall below a specified lower limit, which is necessary to keep the heating coil immersed in water and thus preventing the burning out. This can be achieved by using override control through the lower limit switch (LSS). Under normal circumstances, the selector switch selects the pressure control loop for control; but as soon as the level of water falls below a set value, the selector switch switches to level control mode and the second loop takes over the control action.

Split Range Control
This type of control is used, where there are several manipulated variables, but a single output variable. The coordination among different manipulated variables is carried out by using Split Range Control.

Version 2 EE IIT, Kharagpur 7

Fig. 8 shows an example of a typical split range control scheme. The steam discharges from several boilers are combine at a steam header. Overall steam pressure at the header is to be maintained constant through a pressure control loop. The command from the pressure controller is used for controlling simultaneously the steam flow rates from the boilers in parallel. Clearly, there is a single output variable (steam header pressure) while there are a number of manipulating variables (discharge from different boilers).

References
1. 2. 3. G. Stephanopoulus: Chemical process Control, Prentice Hall of India, New Delhi, 1995. D.R. Coughanowr: Process systems analysis and control (2/e), McgrawHill, NY, 1991. W.L. Luyben and M.L. Luyben: Essentials of Process Control, McgrawHill, NY, 1997.

Version 2 EE IIT, Kharagpur 8

Review Questions

1. Fermentation is a chemical process where sugar, in presence of micro organism breaks down into alcohol and carbon dioxide. Control of fermentation finds wide applications in several industries, e.g. (a) brewing, (b) dairy (c) bakery (d) wine. Fig. Q1 shows a typical control scheme for controlling the dissolved oxygen content in a fermentation process. (a) (b) (c) Explain the operation of the control scheme. What type of control is it? Identify the inner and outer loops in the control system. What types of control would you recommend for the inner and outer loops?

2. Discuss the advantages and disadvantages of using cascade control. 3. Write down the conditions for maintaining stability of a cascade control system. 4. What do you mean by override control? Explain with an example. 5. What do you mean by split range control. Show a schematic arrangement of this type of control.

Version 2 EE IIT, Kharagpur 9

Module 4
Programmable Logic Control Systems
Version 2 EE IIT, Kharagpur 1

Lesson 18
Introduction to Sequence/Logic Control and Programmable Logic Controllers
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
After learning the lesson students should be able to A. Define Sequence and Logic Control B. State three major differences between Logic Control and Analog Control C. Define a Programmable Logic Controller and name its major structural components D. Name the major functions performed by a PLC

What is Sequence and Logic Control?
Many control applications do not involve analog process variables, that is, the ones which can assume a continuous range of values, but instead variables that are set valued, that is they only assume values belonging to a finite set. The simplest examples of such variables are binary variables, that can have either of two possible values, (such as 1 or 0, on or off, open or closed etc.). These control systems operate by turning on and off switches, motors, valves, and other devices in response to operating conditions and as a function of time. Such systems are referred to as sequence/logic control systems. For example, in the operation of transfer lines and automated assembly machines, sequence control is used to coordinate the various actions of the production system (e.g., transfer of parts, changing of the tool, feeding of the metal cutting tool, etc.). Typically the control problem is to cause/ prevent occurrence of ♦ particular values of outputs process variables ♦ particular values of outputs obeying timing restrictions ♦ given sequences of discrete outputs ♦ given orders between various discrete outputs Note that some of these can also be operated using analog control methods. However, in specific applications they may be viewed as discrete control or sensing devices for two reasons, namely, A. The inputs to these devices only belong to two specific sets. For example in the control of a reciprocating conveyor system, analog motor control is not applied. Simple on-off control is adequate. Therefore for this application, the motor-starter actuation system may be considered as discrete. B. Often the control problem considered is supervisory in nature, where the problem is provide different types of supervisory commands to automatic control systems, which in turn carry out analog control tasks, such that over all system operating modes can be maintained and coordinated to achieve system objectives. Examples of some such devices is given below.

Version 2 EE IIT, Kharagpur 3

Industrial Example of Discrete Sensors and Actuators
There are many industrial sensors which provide discrete outputs which may be interpreted as the presence/absence of an object in close proximity, passing of parts on a conveyor, For example, tables x.y and a.b below show a set of typical sensors which provide a discrete set of output corresponding to process variables. Type Signal Remark

Switch

Binary Command Position Set valued Command Temperature Level Position of objects Position of objects Command (unlatched)

External Input Device Feedback Sensor Device External Input Device Feedback Sensor Device Feedback Sensor Device Feedback Sensor Device External Input Device

Limit switch Thumbwheel switch Thermostat Photo cell Proximity detector Push button

Table 3.1 Discrete Sensors

Version 2 EE IIT, Kharagpur 4

Fig. 18.1 Example Industrial Discrete Input and Sensing Devices: Type Output Quantity Energy Source

Relay, Contactor Motor Starter Lamp Solenoid On-off Flow Control valve Directional Valves

voltage motion indication motion Flow Hydraulic Pressure

electrical electrical electrical electrical pneumatic, hydraulic pneumatic, hydraulic

Table 18.2 Example Industrial Discrete Output and Actuation Devices Below we provide an industrial example of Industrial Sequence Control

Point to Ponder: 1
A. Categorise the following sensor systems as Discrete or Continuous a) thermostat b)clinical thermometer c) the infrared sensor in TV sets B. Categorise the following actuator systems as Discrete or Continuous a)the trigger of a gun b) the steering wheel of the car c) a step motor

Version 2 EE IIT, Kharagpur 5

Industrial Example
Piston

Upper limit switch Up Solenoid Down Solenoid Die

Lower limit switch

Fig. 18.2 An Industrial Logic Control Example The die stamping process is shown in figure below. This process consists of a metal stamping die fixed to the end of a piston. The piston is extended to stamp a work piece and retracted to allow the work piece to be removed. The process has 2 actuators: an up solenoid and a down solenoid, which respectively control the hydraulics for the extension and retraction of the stamping piston and die. The process also has 2 sensors: an upper limit switch that indicates when the piston is fully retracted and a lower limit switch that indicates when the piston is fully extended. Lastly, the process has a master switch which is used to start the process and to shut it down. The control computer for the process has 3 inputs (2 from the limit sensors and 1 from the master switch) and controls 2 outputs (1 to each actuator solenoid). The desired control algorithm for the process is simply as follows. When the master switch is turned on the die-stamping piston is to reciprocate between the extended and retracted positions, stamping parts that have been placed in the machine. When the master switch is switched off, the piston is to return to a shutdown configuration with the actuators off and the piston fully retracted.

Point to Ponder: 2
A. Define what is Logic Control system in your own language. Give an example B. In the context of your example show typical objectives in Logic Control

Comparing Logic and Sequence Control with Analog Control
The salient points of difference between Analog Control and Logic/Sequence control are presented in the table below. Version 2 EE IIT, Kharagpur 6

Issue Model

Signal Temporal Property Control On-off/logical Redesign/Tuning Supervisory Open/Closed Loop Infrequent

Logic/Sequence Control Logical State-Transition Simple Model/Easy to build Infrequent Signal range/status (Timed) sequence

Analog Control Numerical Different ial/Difference Eqn Complex Model/Hard to build Liable to change Signal value (Timed)Function/Trajectory linear/non linear analog automatic Open/Closed Loop Tuning needed

Table 18.3 A Comparison of Continuous Variable (Analog) and Discrete Event (Logic/Sequence) Control

Point to Ponder: 3
C. Of Logic Control and Analog Control which one appears simpler and why ? D. Can you cite an example system which requires both Analog and Logic Control?

Programmable Logic Controllers (PLC)
A modern controller device used extensively for sequence control today in transfer lines, robotics, process control, and many other automated systems is the Programmable Logic Controller (PLC). In essence, a PLC is a special purpose industrial microprocessor based realtime computing system, which performs the following functions in the context of industrial operations • • • • Monitor Input/Sensors Execute logic, sequencing, timing, counting functions for Control/Diagnostics Drives Actuators/Indicators Communicates with other computers

Some of the following are advantages of PLCs due to standardized hardware technology, modular design of the PLCs, communication capabilities and improved development program development environment: • • • • Easy to use to simple modular assembly and connection; Modular expansion capacity of the input, outputs and memory; Simple programming environments and the use of standardized task libraries and debugging aids; Communication capability with other programmable controllers and computers

Version 2 EE IIT, Kharagpur 7

Evolution of the PLC
Before the advent of microprocessors, industrial logic and sequence control used to be performed using elaborate control panels containing electromechanical or solid-state relays, contactors and switches, indicator lamps, mechanical or electronic timers and counters etc., all hardwired by complex and elaborate wiring. In fact, for many applications such control panels are used even today. However, the development of microprocessors in the early 1980’s quickly led to the development of the PLCs, which had significant advantages over conventional control panels. Some of these are: • • • • • • • Programming the PLC is easier than wiring physical components; the only wiring required is that of connecting the I/O terminals. The PLC can be reprogrammed using user-friendly programming devices. Controls must be physically rewired. PLCs take up much less space. Installation and maintenance of PLCs is easier, and with present day solid-state technology, reliability is grater. The PLC can be connected to a distributed plant automation system, supervised and monitored. Beyond a certain size and complexity of the process, a PLC-based system compare favorably with control panels. Ability of PLCs to accept digital data in serial, parallel and network modes imply a drastic reduction in plant sensor and actuator wirings, since single cable runs to remote terminal I/O units can be made. Wiring only need to be made locally from that point. Special diagnostic and maintenance modes for quick troubleshooting and servicing, without disrupting plant operations.

However, since it evolved out of relay control panels the PLCs adopted legacy concepts, which were applicable to such panels. To facilitate maintenance and modification of the physically wired control logic, the control panel was systematically organized so that each control formed a rung much like a rung on a ladder. The development of PLCs retained the ladder logic concept where control circuits are defined like rungs on a ladder where each rung begins with one or more inputs and each rung usually ends with only one output. A typical PLC ladder structure is shown below in Fig. 3.2

Version 2 EE IIT, Kharagpur 8

Relays and Contactors
+ve Power busbars -ve Solenoid Lamp Motor coil Relay logic A logical function Realized with relays Fig. 18.3 The structure of Relay Logic Circuits

Relay Ladder
Rung Logic

]
(Virtual) Power Rails • • •

Rung

Relay coil

Fig. 18.4 The structure of Relay Ladder Logic Programs for PLCs

Point to Ponder: 4
E. Name three of the most prominent advantages of the PLCs over hardwired Relay Contactor Logic F. Can you name a single disadvantage in any situation? G. Do you think the idea of developing programs that look like Relay Ladders is very efficient? If so, why? If not, why was it pursued?

Application Areas
Programmable Logic Controllers are suitable for a variety of automation tasks. They provide a simple and economic solution to many automation tasks such as • • • Logic/Sequence control PID control and computing Coordination and communication Version 2 EE IIT, Kharagpur 9

• •

Operator control and monitoring Plant start-up, shut-down

Any manufacturing application that involves controlling repetitive, discrete operations is a potential candidate for PLC usage, e.g. machine tools, automatic assembly equipment, molding and extrusion machinery, textile machinery and automatic test equipment. Some typical industrial areas that widely deploy PLC controls are named in Table x.y. The list is only illustrative and by no means exhaustive. Chemical/ Petrochemical Batch process Pipeline Control Weighing, Mixing Finished Product Handling Water/ Waste Treatment Metals Blast Furnace Continuous Casting Rolling Mills Soaking Pit Steel Melting Shop Manufacturing/Machining Material Conveyors, Cranes Assembly Milling, Grinding, Boring Plating, Welding, Painting Molding/ casting/forming

Table 18.4 Some Industrial Areas for Programmable Controller Applications

Architecture of PLCs
The PLC is essentially a microprocessor-based real-time computing system that often has to handle significant I/O and Communication activities, bit oriented computing, as well as normal floating point arithmetic. A typical set of components that make a PLC System is shown in Fig. 3.5 below.
Remote I/O

MMI

Analog I/O
Memory Digital I/O

Communication Processor Function Modules Network communication

CPU

Backplane

Power

Programmer unit

Printer

Fig. 18.5 Conventional PLC Architecture The components of the PLC subsystem shown in Fig. 3.5 are described below.

Version 2 EE IIT, Kharagpur 10

Central controller
The central controller (CC) contains the modules necessary for the main computing operation of the Programmable controller (PC). The central controller can be equipped with the following: ♦ Memory modules with RAM or EPROM (in the memory sub modules) for the (main memory); ♦ Interface modules for programmers, expansion units, standard peripherals etc; ♦ Communications processors for operator communication and visualization, communication with other systems and configuring of local area networks. A bus connects the CPUs with the other modules. program

Central Processing units
The CPUs are generally microprogrammed processors sometimes capable of handling multiple data width of either 8, 16 or 24 bits. In addition some times additional circuitry, such as for bit processing is provided, since much of the computing involves logical operations involving digital inputs and auxiliary quantities. Memory with battery backup is also provided for the following: ♦ Flags ( internal relays), timers and counters; ♦ Operating system data ♦ Process image for the signal states of binary inputs and outputs. The user program is stored in memory modules. During each program scan, the processor reads the statement in the program memory, executes the corresponding operations. The bit processor, if it exists, executes binary operations. Often multiple central controllers can be configured in hot standby mode, such that if one processor fails the other can immediately pick up the computing tasks without any failure in plant operations.

Communications processors
Communications processors autonomously handle data communication with the following: ♦ Standard peripherals such as printers, keyboards and CRTs, ♦ Supervisory Computer Systems, ♦ Other Programmable controllers, The data required for each communications processors is stored in a RAM or EPROM sub module so that they do not load the processor memories. A local area network can also be configured using communications processors. This enables the connection of various PLCs over a wide distance in various configurations. The network protocols are often proprietary. However, over the last decade, interoperable network protocol standards are also supported in modern PLCs.

Version 2 EE IIT, Kharagpur 11

Program and Data memory
The program and data needed for execution are stored in RAM or EPROM sub modules. These sub modules are plugged into the processors. Additional RAM memory modules can also be connected.

Expansion units
Modules for the input and output of signals are plugged into expansion units. The latter are connected to the central controller via interface modules. Expansion units can be connected in two configurations. A. Centralized configuration The expansion units (EU) are located in the same cabinet as the central controllers or in an adjacent cabinet in the centralized configuration, several expansion units can be connected to one central controller. The length of the cable from the central controller to the most distant expansion unit is often limited based on data transfer speeds. B. Distributed configuration The expansion units can be located at a distance of up to 1000 m from the central controller. In the distributed configuration, up to 16 expansion units can be connected to one central controller. Four additional expansion units can be connected in the centralized configuration to each distributed expansion unit and to the central controller.

Input/Output Units
A host of input and output modules are connected to the PLC bus to exchange data with the processor unit. These can be broadly categorized into Digital Input Modules, Digital Output Modules, Analog Input Modules, Analog Output Modules and Special Purpose Modules. Digital Input Modules The digital inputs modules convert the external binary signals from the process to the internal digital signal level of programmable controllers. Digital Output Modules The digital output modules convert the internal signal levels of the programmable controllers into the binary signal levels required externally by the process. Analog Input Modules The analog input modules convert the analog signals from the process into digital values which are then processed by the programmable controller. Analog Output Modules The analog output modules convert digital values from the programmable controller into the analog signals required by the process. Special Purpose Modules These may include special units for: Version 2 EE IIT, Kharagpur 12

• • • •

High speed counting High accuracy positioning On-line self-optimizing control Multi axis synchronisation, interpolation

These modules contain additional processors, and are used to relieve the main CPU from the high computational loads involved in the corresponding tasks. These are discussed in detail in Lesson 22

Programmers
External programming units can be used to download programs into the program memory of the CPU. The external field programmers provide several software features that facilitate program entry in graphical form. The programmers also provide comprehensive aids for debugging and execution monitoring support logic and sequence control systems. Printer can be connected to the programmers for the purpose of documenting the program. In some cases, special programming packages that run on Personal Computers, can also be used as programming units. There are two ways of entering the program: A. Direct program entry to the program memory (RAM) plugged into the central controller. For this purpose, the programmer is connected to the processor or to the programmer interface modules. B. Programming the EPROM sub modules in the programmer without connecting it to the PC (off-line). The memory sub modules are then plugged into the central controller.

Other Miscellaneous Units
Other units such as Power Supply Units, Bus Units etc. can also be connected to the PLC system.

Point to Ponder: 5
A. Name three major elements of a PLC System B. What is the need for special purpose I/O modules? Explain with an example C. What is a communication Processor?

Version 2 EE IIT, Kharagpur 13

Answers, Remarks and Hints to Points to Ponder Point to Ponder: 1
A. Categorise the following sensor systems as Discrete or Continuous a) thermostat : Discrete b) clinical thermometer : Continuous c) the infrared sensor in TV sets : Discrete B. Categorise the following actuator systems as Discrete or Continuous A) the trigger of a gun: Discrete b) the steering wheel of the car: Continuous c) a step motor: May be considered Discrete or Continuous depending on the mode it is used. If it is used in the incremental mode it may be thought to be discrete (clockwise/anticlockwise). If it is used in the slewing mode, it may be considered continuous

Point to Ponder: 2
A. Define what is Logic Control system in your own language. Give an example Ans: A detailed definition provided in the lesson. The example of a die press is also provided. However, try to give your own example. B. In the context of your example show typical objectives in Logic Control Ans: The following are valid control objectives for the die press example. 1. When MCS is off Up_lamp should never be 1 2. Every transition of the Up_lamp signal from 0 to 1 should be immediately followed by a transition of Dn_lamp from 0 to 1 3. Up_lamp and Dn_lamp can never simultaneously be 1, although they can simultaneously be 0. Such statements are specifications in the sense that the logic controller must ensure that they are satisfied in the controlled system.

Point to Ponder: 3
A. Of Logic Control and Analog Control which one appears simpler and why ? Ans: Analog Control is more complex than logic control. This is because of the fact that logic control models are captured by simple state transition systems containing only a few states. The state space of an analog control system is in infinite. The dynamics can be far more complex than simple state transition systems. Factors such as disturbances must be considered, unlike in logic control B. Can you cite an example system, which requires both Analog and Logic Control? Ans: There are many examples. In a variable air volume air conditioning system, the cooling water temperature is controlled by on-off control of the chiller, while volume of air is Version 2 EE IIT, Kharagpur 14

controlled by analog speed control of the fan. In a CNC machine the speed of the spindle is controlled by analog means, while auxiliaries, such as coolant flow are controlled by PLCs.

Point to Ponder: 4
A. Name three of the most prominent advantages of the PLCs over hardwired Relay Contactor Logic Ans: 1. Programmability 2. Ability to incorporate complex control 3. Expandability, among many others. B. Can you name a single disadvantage in any situation? Ans: For very simple and small systems such as power distribution control a relay based control panel may be a cheaper solution. C. Do you think the idea of developing programs that look like Relay Ladders is very efficient? If so, why? If not, why was it pursued? Ans: It is not efficient. It was pursued, because when PLCs were developed Plant Engineers were more conversant with Relay Logic. So the language was introduced for ease of understanding of Plant Engineers.

Point to Ponder: 5
A. Name three major elements of a PLC System Ans: CPU, I/O Modules, Communication Processor B. What is the need for special purpose I/O modules? Explain with an example Ans: Some i/o operations like high speed counting of shaft encoder pulses to measure speed is very computationally intensive. Therefore to free the CPU from this load, so that other control logics can be computed, special i/o modules with dedicated processors for the task are used. C. What is a communication Processor? Ans: It is a special processor that handles all communication related tasks with other supervisory systems.

Version 2 EE IIT, Kharagpur 15

Module 4
Programmable Logic Control Systems
Version 2 EE IIT, Kharagpur 1

Lesson 19
The Software Environment and Programming of PLCs
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
After learning the lesson students should be able to A. Describe the structure of a PLC Program B. Describe the execution of a PLC Program C. Describe the typical elements of an RLL Diagram D. Design RLL Diagrams for simple industrial logic control problems

Structure of a PLC Program
There are several options in programming a PLC, as discussed earlier. In all the options the common control of them is that PLC programs are structured in their composition. i.e. they consist of individual, separately defined programs sections which are executed in sequence. These programs sections are called ‘blocks”. Each program section contains statements. The blocks are supposed to be functionally independent. Assigning a particular (technical) function to a specific block, which has clearly defined and simple interfaces with other blocks, yields a clear program structure. The testing of such programs in sections is substantially simplified. Various types of blocks are available according to the function of the program section. In general the major part of the program is contained in blocks that contain the program logic graphically represented. For improved modularity, these blocks can be called in a sequence or in nested configurations. Special Function Blocks, which are similar to application library modules, are used to realize either frequently reoccurring or extremely complex functions. The function block can be “parameterized”. The interface to the operating system of the PLC, which are similar to the system calls in application programming for Personal Computers, are defined in special blocks. They are only called upon by the system program for particular modes of execution and in the case of the faults. Function blocks are also used where the realization of the logic control STEP 5 statements can’t be carried out graphically. Similarly, individual steps of a control sequence can be programmed into such a block and reused at various points in a program or by various programs. PLC manufacturers offer standard functions blocks for complex functions, already tested and documented. With adequate expertise the user can produce his own function blocks. Some very common function blocks (analog input put, interface function blocks for communication processors and others) may be integrated as standard function blocks and supported by the operating system of the PLC. Users can also define separate data blocks for special purposes, such as monitoring, trending etc., and perform read/write on such areas.

Version 2 EE IIT, Kharagpur 3

Such facilities of structured programming result in programs, which are easier to read, write, debug and maintain.

Program Execution
There are different ways and means of executing a user program. Normally a cyclic execution program is preferred and this cyclic operators are given due priorities. Program processing in a PLC happens cyclically with the following execution: 1. After the PLC is initialised, the processor reads the individual inputs. This status of the input is stored in the process- image input table (PII). 2. This processor processes the program stored in the program memory. This consists of a list of logic functions and instructions, which are successively processed, so that the required input information will already be accessed before the read in PII and the matching results are written into a process-image output table (PIQ). Also other storage areas for counters, timers and memory bits will be accessed during program processing by the processor if necessary. 3. In the third step after the processing of the user program, the status from the PIQ will transfer to the outputs and then be switched on and/or off. Afterwards it begins the execution of the next cycle from step 1. The same cyclic process also acts upon an RLL program. The time required by the microprocessor to complete one cycle is known as the scan time. After all rungs have been tested, the PLC then starts over again with the first rung. Of course the scan time for a particular processor is a function of the processor speed, the number of rungs, and the complexity of each rung. Initialise

Read all inputs

Program

Update all

outputs

Fig. 19.1 The cyclic execution of PLC Programs Version 2 EE IIT, Kharagpur 4

Interrupt Driven and Clock Driven Execution Modes
A cyclically executing program can however be interrupted by a suitably defined signal resulting in an interrupt driven mode of program execution (when fast reaction time is required). If the interrupting signal occurs at fixed intervals we can also realized time synchronous execution (i.e. with closed loop control function). The cyclic execution, synchronized by a real time clock is the most common program structure for a PLC. Similarly, programmers can also define error-handling routines in their programs. Specific and defined error procedures are then invoked if the PLC operating system encounters fault of given types during execution.

Point to Ponder: 1
A. What are the different modes of execution? B. Which is the most common? C. State for each of the others, when these are to be used. D. Give examples for your arguments if you can

Programming Languages
PLC programs can be constructed using various methods of representation. Some of the common ones are, described below.

The Relay Ladder Logic (RLL) Diagram
A Relay Ladder Logic (RLL) diagram, also referred to as a Ladder diagram is a visual and logical method of displaying the control logic which, based on the inputs determine the outputs of the program. The ladder is made up of a series of “rungs” of logical expressions expressed graphically as series and parallel circuits of relay logic elements such as contacts, timers etc. Each rung consist of a set of inputs on the left end of the rung and a single output at the right end of each rung. The structure of a rung is shown below in Fig. 19.1(a) & (b). Fig. 19.1 shows the internal structure of a simple rung in terms its element contacts connected in a series parallel circuit.

Version 2 EE IIT, Kharagpur 5

Rung Logic

]
(Virtual) Power Rails • •

Rung Relay coil

Fig. 19.1(a) The structure of Relay Ladder Logic Programs for PLCs

Fig. 19.1(b) The internal structure of a simple Rung

RLL Programming Paradigms: Merits and Demerits
For the programs of small PLC systems, RLL programming technique has been regarded as the best choice because a programmer can understand the relations of the contacts and coils intuitively. Additionally, a maintenance engineer can easily monitor the operation of the RLL program on its graphical representation because most PLC manufacturers provide an animated display that clearly identifies the states of the contacts and coils. Although RLL is still an important language of IEC 1131-3, as the memory size of today's PLC systems increases, a large-sized RLL program brings some significant problems because RLL is not particularly suitable for the well-structured programming: It is difficult to structure an RLL program hierarchically.

Version 2 EE IIT, Kharagpur 6

Example: Forward Reverse Control
IN002 IN001 OP001 OP002 OP001

IN001

IN003 OP002 OP001

OP002

Fig. 19.2 RLL Diagram for the Forward Reverse Control Problem This example explains the control process of moving a motor either in the forward direction or in the reverse direction. The direction of the motor depends on the polarity of the supply. So in order to control the motor, either in the forward direction or in the reverse direction, we have to provide the supply with the corresponding polarity. The Fig 19.2 depicts the procedure to achieve this using Relay Ladder Logic. Here, the Ladder consists of two rungs corresponding to forward and reverse motions. The rung corresponding to forward motion consists of 1. A normally closed stop push-button (IN001), 2. A normally opened forward run push-button (IN002) in parallel with a normally opened auxillary contact(OP001), 3. A normally closed auxillary contact(OP002) and 4. The contacter for coil(OP001). Similarly, the rung corresponding to reverse motion consists of 1. A normally closed stop push-button (IN001), 2. A normally opened forward run push-button (IN003) in parallel with a normally opened auxillary contact(OP002), 3. A normally closed auxillary contact(OP001) and 4. The contacter for coil(OP002). Operation: The push-buttons(PB) represented by IN--- are real input push-buttons, which are to be manually operated. The auxillary contacts are operated through program. Initially the machine is at standstill, no voltage supply is present in the coils, and the PBs are as shown in the fig. The stop PB is intially closed, the motor will not move until the forward run PB/reverse run PB is closed. Suppose we want to run the motor in the forward direction from standstill, the outputs of the coils contacters have logic ‘0’ and hence both the auxillary contacts are turned on. If we press and release the forward run PB, the positive voltage from the +ve voltage rail is passed to the Version 2 EE IIT, Kharagpur 7

coil. Once the coil contacter gives the logic ‘1’, the following consequences takes place simultaneously A. The auxillary contact OP001 in the second rung becomes opened,which stops the voltage for reverse motion of the motor. At this stage, the second rung is not turned on even the reverse run PB is pressed by mistake. B. The auxillary contact OP001 is the first rung is on, which provides the path for the positive voltage until the stop PB is pressed. Here the auxillary contact OP001 acts as a ‘latch’, which facilitates even to remove the PB IN002 once the coil OP001 is on. If we want to rotate the motor in the reverse direction, the stop PB is to be pressed sothat no voltage in the coil is present, then we can turn on the PB corresponding to reverse run. This is a simple example of ‘interlocking’, where each rung locks the operation of the other rung. There are several other programming paradigms for PLCs. Two of them are mentioned here for briefly. The Function Chart (IEC) Depicts the logic control task symbols in terms of functional blocks connected symbolically in a graphic format. The Statement List (STL) Is made up of series of assembly language like statements each one of which represents a logic control statement executable by the processor of the programmable controller. The statement list is the most unrestricted of all the methods of representation. Individual statements are made up of mnemonics, which represent the function to be executed. This method of representation is favoured by those who have already had experience in programming microprocessors or computers.

Point to Ponder
Can you think of a control logic which would be difficult to program in the RLL framework?

Typical Operands of PLC Programs
In the RLL Program that we have already seen we have already encountered contacts and output coils. However, there are some other elements of a RLL diagram which are commonly used in industrial applications. These are described in some detail below.

Inputs I, Output Q
Generally, the operands of a PLC program can be classified as inputs (I) and outputs (Q). The input operands refer to external signals of the controlled system, whose values are acquired from the input signal modules. The operating system of the PLC assigns the signal status of the input and output modules into the process image of the inputs and outputs at the beginning of the program. The operand area is located within the process image of the central controller RAM. Within the program, the signal status of the operand area is scanned and processed into logic functions in accordance with the user program; the individual bits within the process output image are fixed. When the program has been executed the operating system transfers the signal Version 2 EE IIT, Kharagpur 8

status of the process image independently to the output modules. This method enables faster program execution because access to the process image is executed much faster than access to the I/O – modules. In RLL Programs, inputs are represented as contacts. Two types of contacts are used, namely, normally open and normally closed contacts. The difference in the sense of interpretation between these contacts is shown below. The switch shown here is a NO contact, i.e. it is closed when it is active.

NO contact inactive

NO contact open

NO contact active

NO contact closed

Fig. 19.3 (a) NO Contact interpretation The switch shown here is a NC contact. i.e. it is closed when it is not active.

NC contact inactive

NC contact closed

NC contact active

NC contact open

Fig. 19.3 (b) NC Contact interpretation

Internal Variable Operands or Flags
In addition to the inputs and outputs, which correspond to physical signals in the controlled systems internal variables are required to save the intermediate computational values of the program. These are referred to as Flags, or the Auxiliary Contacts in Relay Ladder Logic parlance. The number of such variables admissible in a program may be limited. Such auxiliary contacts correspond to output values and are assumed to be activated by the corresponding output values. They may be either of an NO or an NC type. Therefore, an NO auxiliary contact would be closed if the corresponding output is active i.e. has value “1”.

Point to Ponder: 2
A. What is the basic difference between Input and Auxiliary Contacts? C. Design an RLL Program for the following Industrial Problem In a controlled plant three fans are to be monitored. If at least two fans are running, the indicator light of the monitor is permanently switched on. The indicating lamp blinks slowly if only one fan is running (with 0.5 Hz) and rapidly (with 2 Hz) if no fan at all is on. the monitor is only active when the signal “plant in operation” signal status “1” is activated. Otherwise the indicating light is switched off. The function “lamp test” can be Version 2 EE IIT, Kharagpur 9

carried out with the signal “plant in operation”. At signal status “1” of this signal the indicating light is either permanently on or is flashing.

Timer
These are special operands of a PLC, which represent a time delay relay in a relay logic system. The time functions are a fixed component of the central processing unit. The number of these varies from manufacturer to manufacturer and from product to product. It is possible to achieve time delays in the range of few milliseconds to few hours. Set time delay value Preset Reg Timing Reg Enable Reset Logic Output coil Goes high When TR = PR

Run Logic

Stores Current Time

Fig. 19.4 Structure of a Typical Timer Representation for timers is shown in Fig. 19.4. Timers have a preset register value, which represent the maximum count it can hold and can be set using software/program. The figure shown below has a ‘enable reset logic’ and ‘run logic’ in connection with the timer. The counter doesnot work and the register consists of ‘zero’ until the enable reset logic is ‘on’. Once the ‘enable reset logic’ is ‘on’, the counter starts counting when the ‘run logic’ is ‘on’. The output is ‘on’ only when the counter reaches the maximum count. Various kinds of timers are explained as follows On delay timer: The input and output signals of the on delay timer are as shown in the Fig. 19.6. When the input signal becomes on, the output signal becomes on with certain delay. But when the input signal becomes off, the output signal also becomes off at the same instant. If the input becomes on and off with the time which less than the delay time, there is no change in the output and remains in the ‘off’ condition even the input is turned on and off i.e., output is not observed until the input pulse width is greater than the delay time.

Version 2 EE IIT, Kharagpur 10

Input

Output Delay

Fig. 19.5 A Typical Input output waveform for an On-Delay Timer

I01

O01

O01

PR= Delay Value

O02

O01 Fig. 19.6 The realization of an On-delay timer from a general timer. Realization of on-delay timer: The realization of on-delay timer using the basic timer shown in the previous fig is explained here. The realization is as shown in the Fig. 19.6, which shows a real input switch(IN001), coil1(OP002), two normally opened auxillary contacts(OP002), coil2(OP002). When the real input switch is ‘on’ the coil(OP002) is ‘on’ and hence both the auxillary switches are ‘on’. Now the counter value starts increasing and the output of the timer is ‘on’ only after it reaches the maximum preset count. The behaviour of this timer is shown in figure, which shows the on-delay timer. The value in the counter is ‘reset’ when the input switch(IN001) is off as the ‘enable reset logic’ is ‘off’. This is a non-retentive timer. Off delay timer: The input and output signals of the off delay timer are as shown in the Fig. 19.7. When the input signal becomes on, the output signal becomes on at the same time. But when the input signal becomes off, the output signal becomes ‘off’ with certain delay. If the input becomes on and off with the time which less than the delay time, there is no change in the output and remains in the ‘on’ condition even the ipnut is turned on and off i.e., the delay in the output is not observed until the input pulse width is greater than the delay time.

Version 2 EE IIT, Kharagpur 11

Input

Output

Delay

Delay Timer Starts

Del But resets

Fig. 19.7 A Typical Input output waveform for an Off-Delay Timer Realization of off-delay timer: The realization of on-delay timer using the basic timer shown in the previous fig is explained here. The realization is as shown in the Fig. 19.8, which shows a real input switch(IN001), coil1(OP002), two normally closed input contacts(IN001), output contacts (OP002,OP003). When the real input switch is ‘on’, the coil(OP002) is ‘on’ and both the auxillary input switches are ‘off’. Now the output contact(OP002) becomes ‘off’ which in turn makes the auxillary contact(OP002) in the third rung to become ‘on’ and hence the output contact(OP003) is ‘on’. When the real input switch is ‘off’, the counter value starts increasing and the output of the contact becomes ‘on’ after the timer reaches the maximum preset count. At this time the auxillary contact in the third rung becomes ‘off’ and so is the output contact(OP003). The input and output signals are as shown in the figure, which explain the offdelay timer.

I01

O01

I01

PR = delay value

O02

I01 I01 O02

O03 O03 Fig. 19.8 The realization of an Off-delay timer from a general timer. Fixed pulse width timer: The input and output signals of the fixed pulse width timer are as shown in the Fig 19.9. When the input signal becomes on, the output signal becomes on at the Version 2 EE IIT, Kharagpur 12

same time and remains on for a fixed time then becomes ‘off’. The output pulse width is independent of input pulse width.

Input

Output Fig. 19.9 A Typical Input output waveform for a Fixed Width Timer

I01

O01

I01 I01

PR= Pulse width

O02

I01

O02

O03

Fig. 19.10 The realization of a Fixed width timer from a general timer. Retentive timer: The input and output signals of the retentive timer are as shown in the 19.11. This is also implemented internally in a register as in the previous case. When the input is ‘on’ , the internal counter starts counting until the input is ‘off’ and at this time, the counter holds the value till next input pulse is applied and then starts counting starting with the value existing in the register. Hence it is named as ‘retentive’ timer. The output is ‘on’ only when the counter reaches its ‘terminal count’.

Version 2 EE IIT, Kharagpur 13

Input T ACC A B A+B+C=T Reset Fig. 19.11 A Typical Input output waveform for a Retentive Timer Non-retentive timer: The input and output signals of the non-retentive timer are as shown in the Fig. 19.11. This is implemented internally in a register. When the input is ‘on’, the internal counter starts counting until the input is ‘off’ and at this time the value in the counter is reset to zero. Hence it is named as non-retentive timer. The output is ‘on’ only when the counter reaches its ‘terminal count’. C

Input Preset Output

Fig. 19.12 A Typical Input output waveform for a Fixed Width Timer

Point to Ponder: 3
Draw the timing diagrams for the output coils in the RLL realizations of an on-delay, an offdelay and a fixed pulse width timer

Counter
The counting functions (C) operate as hardware counters, but are a fixed component of the central processing unit. The number of these varies for each of the programmable controllers. It is possible to count up as well as to count down. The counting range is from 0 to 999. The count is either dual or BCD coded for further processing. Version 2 EE IIT, Kharagpur 14

Counter
Set terminal Count valve Count Logic CR incremented by 1 every time Count logic goes high Preset Count Enable/Reset Logic Output coil Stores Current count Goes high When CR = PR

Fig. 19.13 Structure of a Typical Counter

User defined Data
If the memory capacity of the flag area is not sufficient to memorize the signal status and data, the operand area “data” (D) is applied. In general, in the flag area, primarily binary conditions apply, whereas in the data area digital values prevail and are committed to memory. The data is organized into data blocks (DB). 256 data words with 16 bit each can be addressed to each data block. The data is stored in the user memory sub module. The available capacity within the module has to be shared with the user program.

Addressing
The designation of a certain input or output within the program is referred to as addressing. Different PLC manufacturers adopt different conventions for specifying the address of a specific input or output signal. A typical addressing scheme adopted in PLCs manufacturers by Siemens is illustrated in the sequel. The inputs and outputs of the PLCs are mostly defined in groups of eight on digital input and/or digital output devices. This eight unit is called a byte. Every such group receives a number as a byte address. Each in/output byte is divided into 8 individual bits, through which it can respond with. These bits are numbered from bit 0 to bit 7. Thus one receives a bit address. For example, in the address I0.4, I denotes that the address type is specified as Input, 0 is the byte address and 4 the bit address. Similarly in the address Q5.7, Q denotes that the address type is specified as Output, 5 is the byte address and 7 is the bit address.

Operation Set
The operation set of PLC programming languages can be divided into four major function groups: Binary or Logic functions Version 2 EE IIT, Kharagpur 15

Numeric or Arithmetic functions Program control functions and Other statements Binary function combines primarily binary signal status with logic operations. The logic functions (binary logic operation) are the AND and the OR functions, according to the series and the parallel circuit arrangement on the ladder diagram. The result of the logic operation together with the memory function is then assigned to the appropriate operand. The majority of operands are inputs (I), output (O) and flags (F). With the result of logic operations of binary logic, timers can be enabled and started and counters can be initiated, incremented to count up or decremented to count down. Because the results of the time and count functions can be combined with logic functions, the associated operations are considered to be part of the group of binary functions. Arithmetic functions are primarily used to perform arithmetic on numerical values. These can be combined with logic operations, such that the numeric operations can be enabled or disabled in a given scan based on logic conditions, much like “if then else” programming constructs. Similarly, logic conditions can be derived from numeric variables using operations like comparison. Logic operations are established in the register of the processor (in the “accumulators”). The registers are loaded with a loading operation (they are supplied with a value). The result of the logic operation is then transferred back to the operand area via a transfer operation. Digital functions are: Program Control operations include, Function block calls and Jump functions.

Answers, Remarks and Hints to Points to Ponder Point to Ponder: 1
E. What are the different modes of execution? Ans: Cyclic, Interrupt Driven and Clock Driven F. Which is the most common? Ans: The cyclic Mode is most common G. State for each of the others, when these are to be used. Ans: The interrupt driven mode is to be used for those tasks that need an immediate response during any time of normal execution, but as such are sporadic. Clock driven modes are used for those tasks which have to precisely synchronized with time relative to some defined clocks. H. Give examples for your arguments if you can Ans: Emergency Shutdown/Alarm Tasks are programmed using interrupt driven modes.

Version 2 EE IIT, Kharagpur 16

Communication tasks to supervisory computers for say trend updates can be implemented with clock-driven tasks.

Point to Ponder: 2
A. What is the basic difference between Input and Auxiliary Contacts? Ans: Input contacts correspond to physical devices that can be asserted either by external agents or by the process itself for feedback. Auxiliary contacts are basically memory locations storing intermediate logical results and do not correspond to physical devices. B. Design example of Fan monitor Ans: In a controlled plant three fans are to be monitored. If at least two fans are running, the indicator light of the monitor is permanently switched on. The indicating lamp blinks slowly if only one fan is running (with 0.5 Hz) and rapidly (with 2 Hz) if no fan at all is on. the monitor is only active when the signal “plant in operation” signal status “1” is activated. Otherwise the indicating light is switched off. The function “lamp test” can be carried out with the signal “plant in operation”. At signal status “1” of this signal the indicating light is either permanently on or is flashing. Hints: The total logic control is primarily made up of 4 elements Scan, if at least two fans are funning. Scan, if no fan is running Scan, if only one fan is running Summary of all three scans and logic control with the signal “plant in operation”

Point to Ponder: 3
Ans:

On delay timer On – Delay Timer

O01

O02 On delay

Version 2 EE IIT, Kharagpur 17

Off Delay timer
OP001

OP002

OP003 Off Delay

Fixed Pulse Width Fixed Pulse width
I01

003

Version 2 EE IIT, Kharagpur 18

Module 4
Programmable Logic Control Systems
Version 2 EE IIT, Kharagpur 1

Lesson 20
Formal Modelling of Sequence Control Specifications and Structured RLL Programming
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
After learning the lesson students should be able to A. Describe motivations for formal modelling in the design of sequence control programs for an industrial control problem. B. Describe the major steps in the design of a sequence control program for an industrial control problem. C. Develop a Finite State machine model for simple industrial control problems D. Develop a sequence control program for a Finite State Machine model

Motivation for Formal Modelling
To be convinced about the need for developing formal models for control problems in a systematic way, consider the example shown below.

Industrial Logic Control Example Revisited
Consider the industrial logic control example of the stamping process presented in Lesson 18. Let us recall the description of the process given in the lesson. For convenience of reference the description and a pictorial representation of the process is reproduced below. Piston

Up Solenoid Down Solenoid

Top limit switch Die Bottom limit switch

Fig 20.1 An Industrial Logic Control Example

Linguistic description of the industrial stamping process
The die stamping process is shown in figure 20.1. This process consists of a metal stamping die fixed to the end of a piston. The piston is extended to stamp a work piece and retracted to allow the work piece to be removed. The process has 2 actuators: an up solenoid and a down solenoid, which respectively control the electro-hydraulic direction control valves for the extension and retraction of the stamping piston and die. The process also has 2 sensors: an upper limit switch Version 2 EE IIT, Kharagpur 3

that indicates when the piston is fully retracted and a lower limit switch that indicates when the piston is fully extended. Lastly, the process has a master switch which is used to start the process and to shut it down. The control computer for the process has 3 inputs (2 from the limit sensors and 1 from the master switch) and controls 2 outputs (1 to each actuator solenoid). The desired control algorithm for the process is simply as follows. When the master switch is turned on, the die-stamping piston is to reciprocate between the extended and retracted positions, stamping parts that have been placed in the extended piston machine. When the master switch is switched off, the piston is to return to a shutdown configuration with the actuators off and the piston fully retracted. At first, let us consider an Relay Ladder Logic program that has been written directly from the linguistic description and assess it for suitability of operations.

The first version of sequence control program for the industrial stamping process
Master Switch Bottom LS Up_sol

Dn_sol Up_sol Master Switch Top_LS Top_LS Dn_sol

Up_sol Dn_sol Bottom LS

Fig. 20.2 An RLL program for the industrial stamping process A hastily constructed RLL program for the above process may look like the one given in Figure 20.2. The above program logic indicates that the Up solenoid output becomes activated when the Master Switch is on and the bottom Limit Switch is on. Also there is interlock provided, so that when the Down solenoid is on, the Up solenoid cannot be on. Further, once the Up solenoid is on, the output is latched by an auxiliary contact, so that it remains on till the bottom LS is made on, when it turns off. A similar logic has been implemented for the activation of the Down solenoid. However, on closer examination, several problems may be discovered with the above program. Some of these are discussed below. • For example, there is no provision for a Master stop switch to stop the press from stopping in an emergency, except by turning the Master Switch off. This would indeed stop the process, however, if the press stops midway, both bottom and top limit switches would be off. Now the process would not start, even if the Master switch is turned on again. Therefore, either a manual jogging control needs to be provided, so that the Version 2 EE IIT, Kharagpur 4

operator can return the piston to the up position by manually operating the hydraulics, or special auto mode logic should be designed to perform this. • As a second example, note that this process does not have a part detect sensor. This implies that the moment the Master switch is on the press would start going up and down at its own travel speed, regardless of whether a part has been placed for pressing or not. Apart from wastage of energy, this could be safety hazard for an operator who has to place the part on the machine between the interval of a cycle of operation.

The above discussion clearly indicates the need for a systematic approach towards the development of RLL programs for industrial logic control problems. This is all the more true since industrial process control is critical application domain where control errors can lead to loss of production or operator safety. Therefore, in this chapter, we discuss a systematic approach towards the design of RLL programs.

Point to Ponder: 1
A. Before going through the rest of the system description attempt to modify the RLL program shown in Fig. 20.2 to ensure that the process always moves to the up position first, and then from there, resumes its press cycle B. Before going through the rest of the system description attempt to modify the RLL program shown in Fig. 20.2 to include a part detect sensor C. Attempt to merge the above two solutions so that the new program is free from both the above problems

Steps in Sequence Control Design
The approach to Sequence control Design presented below is derived from the basic principles of modern software engineering practices. An interested student is referred to standard software engineering references for a more detailed discussion on these. A brief description of the steps follows: The broad initial steps are - Requirement Analysis - For Modelling of the process - Design - Design verification and validation

A. Requirements Analysis
This is a very important initial step. Errors in this step may be discovered very late during commissioning of the control system and can result in loss of significant amount of man-hours. Note that, since the control engineer is not necessarily an expert in plant operations. Therefore it is quite natural that he may not understand requirements and characteristics of process operations fully, without conscious and significant effort.

Version 2 EE IIT, Kharagpur 5

Therefore, this phase is to be carried out in close consultation with the plant engineers of the user organization. It basically involves studying the system behavioral aspects of the system to: i) identify the feedback inputs from sensors and the external operator inputs from Man Machine Interface (MMI), ii) identify the controller outputs to actuators and the outputs to the indicators/ MMI, iii) study and document the sequence of actions and events under the various operational ‘modes’. This should not only indicate the ‘normal’ modes, but also ‘emergency’, ‘startup’, ‘shut-down’ and other special modes operation that may not be occurring very frequently but important all the same. iv) study the effects of possible failures in the process as well as sensors and identify possible means of recovery. Although a failure may be very rare in occurrence, unless they are considered, some of them may have devastating consequences when they occur. Failures in a process can occur in the sensors, the actuators, or some time the process equipment itself. v) examine need for manual override control, additional sensors, indicators, and alarms for maintenance, operational efficiency or safety. Initially, often the above information may be collected in the form of informal linguistic descriptions by direct discussion with plant personnel. Further, it must be remembered that software development is often an iterative process and therefore, the above analysis steps may have to be repeated a number of times for an actual design exercise.

Requirements Analysis for the Stamping Process
As given above the first step in requirement analysis is to identify the Process Control Inputs that need to be sensed and the Process Control outputs that need to be actuated.

Process Control Inputs
• Part sensor: A position switch that detects when a part has been placed. In cases where proper positioning of the part can take time. One may also consider using manual switch to be operated by the operator once he is satisfied that the part is properly placed and ready to be stamped. Here an automated part detect sensor has been assumed. Auto PB : A push button that indicates that machine is ready to stamp parts one after the other in the ‘automatic mode’ Stop PB: A push button that the operator can use to stop the machine any time during the time that the piston is moving down. This is needed to avoid stamping a part if any last second error is discovered by the operator regarding, say, the placement of the part. Reset PB: In case the piston has been stopped due to some error condition, it is desired that the operator explicitly presses this push button to indicate that the error has been taken care of, and the machine is ready to return to stamping in the auto mode. Bottom LS: This sensor indicates when the piston has reached the bottom position. Top LS: This sensor indicates when the piston has reached the top position. Version 2 EE IIT, Kharagpur 6

• •

• •

Process Control Outputs
• • • • Up Solenoid: Control output that drives the Up solenoid of the electro-hydraulic direction control valve which in turn drives the piston up. Down Solenoid: Control output that drives the Up solenoid of the electro-hydraulic direction control valve which in turn drives the piston up. Auto Mode Indicator: An indicator lamp that indicates that the machine is in ‘Auto’ mode. Part Hold: A gripping actuator that holds the part firmly to avoid movements during stamping

Point to Ponder: 2
A. Can we use corresponding toggle switches, instead for the Auto, Stop and Reset PBs? What difficulties may be encountered? B. For the limit switches, and the part detect sensor, would you prefer mechanical switches over photo switches for this application? Justify. C. Propose at least one additional each of sensors, indicators and actuators for the above application and mention their benefits.

Sequence of Events and Actions
A. The “Auto” PB turns the Auto Indicator Lamp on B. When a part is detected, the press ram advances down to the bottom limit switch C. The press then retracts up to top limit switch and stops D. A “Stop” PB, if pressed, stops the press only when it is going down E. If the “Stop” PB is pressed, the “Reset” PB must be pressed before the “Auto” PB can be pressed F. After retracting, the press waits till the part is removed and the next part is detected

Point to Ponder: 3
A. Note that in step F above, it is important to detect that the part is removed. What would happen, if this is not detected? B. What would happen if after Stop PB is pressed, Reset PB and Auto PB are pressed in that sequence, even if the piston has not been taken to the top position manually?

Effects of failures
Among possible failures for this process are drops in hydraulic pump pressure, failures in top and bottom limit switches etc. The exact nature of the failures and its impact need to be understood for the application context. While this should be done, it requires domain knowledge for the Version 2 EE IIT, Kharagpur 7

control engineer. This is therefore not attempted here for reasons of conciseness. However, the learner is encouraged to augment the control logic with additional logic to detect such failures rapidly and initiate activities for fault tolerance.

Point to Ponder: 4
A. What would happen in the process controlled by the program shown in Fig. 20.2 if top/bottom Limit switch is stuck closed/open? B. What would happen in the process controlled by the program shown in Fig. 20.2 if the hydraulic pump pressure becomes too low to move the RAM?

Formal process modelling
Once the requirements have been ascertained, formal process modelling can be undertaken. In this step the informal linguistic descriptions have to be rigorously checked for ambiguity, inconsistency or incompleteness. This is best achieved by converting linguistic descriptions into formal process models. Initially one may use intermediate forms like list of operations, flowchart etc. Eventually and before developing the control programs, these are to be converted into mathematically unambiguous and consistent description using a formal modeling framework such as a Finite State Machine (FSM). It is the experience of practical engineers that modelling paradigms that can be represented pictorially are particularly suited to human beings. For formal modelling, a process often can be viewed as a Discrete Event System (DES). Many formalisms for creating timed or untimed models of DESs exist (e.g. Petri Nets). A detailed description of these is beyond the scope of this lesson. An interested reader is referred to literature on real-time systems for a more detailed discussion on these. In this lesson, it is shown how the process dynamics can be modeled as a Finite State Machine. The following facts which are very important to modelling are mentioned. A. An FSM is a simple formalism for DES in which, at any time, the system exists in any one discrete-state of a finite set of such states. B. A state is basically an assignment of values to the set of variables of the system. For a discrete event system, the process variables are assumed to take only a finite set of values. For example, the limit switches can only take two values each, namely, either ON or OFF. C. Further, the set of the process variables have to be chosen in such a manner that, the future behaviour of the process would be determined solely based on the values of the chosen set of variables at the present time. For example, for the stamping press example, the set of process variables would include the values for the Top and Bottom limit switches. However, based only on these the behaviour of the process cannot be determined. This is because from these it cannot be determined whether the piston is moving up or moving down. Therefore one would have to add the state of the motion as a state variable. The set of values that this variable can take are: ‘going up’, ‘going down’ and ‘stationary’. In this case, one would also have to add another variable, namely the value of the part detect sensor output to be able to distinguish between the behavioral difference between the case when it is ON and when it is OFF, when the piston is at the top position. D. Some of the state variables may be measured physically with sensors. Others may not be.

Version 2 EE IIT, Kharagpur 8

E. The choice of state variables can be subjective and different designers might pick others. The choice also depends on the nature of control actions that one would like to take. Thus, the choice of states is specific to the machine and its operation. F. During its life cycle, the process moves from state to state over time. Thus it spends most of its time in the states. Occasionally however, it makes a transition from one state to another. The occurrence of a transition depends entirely on the occurrence of discrete events. Such events are names given to conditions involving states, some of which may change due to external factors, such as operator inputs, or due to internal factors, such passage of time. On occurrence of such an event, mechanisms causing state transitions are triggered. State transitions or events are generally considered instantaneous and thus, the system spends time only in the various states. State variables are modified by the occurrence of transitions. In fact, it is this change in the values of the state variables, which is taken to be a transition from one state to another. G. All possible combinations of state variables may not be valid state assignments for a system. In other words, the system can have only some of all the possible combinations of state variables. These are said to be the combinations that are ‘reachable’ by the system. H. One of the states is generally taken to be an ‘initial state’. The system, when it starts its life cycle, that is, at the time from which its behaviour is described by the State Diagram, is supposed to be at the initial state. I. At each state, a set of outputs are exercised. This is described by an output table, where the values for each output variable at each of the states is shown. The output table for the stamping press is shown in Figure 20.x. 1 A 2 B 3 C D 5 6 4 E F

G

Figure 20.3 The State Diagram for the industrial stamping press

Version 2 EE IIT, Kharagpur 9

State No. 1 2 3 4 5 0 0 0 0 O/P Auto Indicator Part Hold Up Sol Down Sol 0 0 0 0 1 0 0 0 1 1 0 1 1 1 1 0 1 0 0 0 6

Figure 20.4 The Output Table for the industrial stamping press From the State Diagram in Fig. 20.3, the following may be noted. A. The states are marked numerically using the numerals 1-6. At each state, one has a unique set of values for the process state variables. For example, in state 3, both the limit switches are OFF, the part detect switch in ON, the motion state variable of the piston is ‘going down’. State 1 is marked as the initial state with a double square. B. The transitions are marked alphabetically using the letters A-G. Each transition has an associated condition under which it occurs. For example, the condition for transition B may be simply stated as “when it in state 2 and part detect goes ON’. Note that the condition described by the phrase ‘in state 2’ can be further explained in terms of the values of the state variable corresponding to state 2. C. The outputs exercised at each state are described by the Output Table in Fig. 20.4.

Design of RLL Program
Based on the formal model, the sequence control program can be developed systematically. In fact, one of the main advantages of formal process modelling is that the process of development of the control program becomes mechanical. Thus it can be done quickly and with a much reduced chance of error. In the case of FSM models one has to write the RLL program such that over the scan cycles it executes the state machine itself. The outputs of the state machine go to the process, and since the state machine is nothing but a behavioral model for the process, the process also executes the transitions of the machine, as desired. The realization of a state machine by an RLL program involves computation of the process transitions, in terms of the inputs and the internal state variables of the program followed by computation of the new states and finally, the outputs corresponding to the states. This method is demonstrated here for RLL programming using the above example of the industrial stamping process. The ladder logic begins with a section to initialize the states and transitions to a single value, corresponding to the initial state. Some PLCs programming languages provide special instructions for such initialization. In this case, however, it is assumed that all auxiliary variables representing the states are set to zero initially. Logic is provided such that in the first scan the auxiliary state variable corresponding to the initial state would be set to 1. The next section of the ladder logic considers the transitions and then checks for transition conditions. Each transition condition contains an auxiliary NO contact corresponding to the source state from which it is defined. For example, note that the logic for the rung corresponding to transition A contains a contact corresponding to state 1, which is the source state for transition A. Further, it contains the other logical terms corresponding to the input state variables as well as

Version 2 EE IIT, Kharagpur 10

timer outputs, if applicable. In the case of transition A, the external condition is simply the pressing of the Auto PB. Note that, in any scan cycle, at most one transition can be enabled. The next block of rungs constitutes the state logic. If the transition logic for any transition is satisfied, the following state logic which is the destination state for the enabled transition is to be turned on and the state logic which is the source state for the enabled transition is to be turned off. Therefore each of the rungs corresponding to a state contains one auxiliary NO contact corresponding to the transition for which that state is a destination state. Similarly, each of the rungs corresponding to a state contains one auxiliary NC contact corresponding to the transition for which that state is a source state in series with the above NO contact. If there is more than one transition for which the state is a destination, all the auxiliary NO contacts corresponding to these transitions should be put in parallel. Similarly, if there is more than one transition, for which the present state is a source, all the auxiliary NC contacts corresponding to these transitions are to be put in series. This occurs in the same scan cycle in which the transition logic is turned on, since the state logic rungs follow those corresponding to transition logic. Note that at the end of every scan cycle, there is only one state logic that is enabled. Now that the state logic has changed, in the next scan cycle the transition that was enabled, turns off and the system stays in that state, till the next transition logic gets enabled. So that the state logic remains turned on, even if a transition for which this state is a source, turns off, an NO auxiliary contact corresponding to the state that latches the state logic is to be provided in parallel with the parallel block of all the auxiliary NO contacts for each transition for which the present state is a destination. State 1 Auto PB Transition A Transition B

State 2

Part Detect

State 3

Bottom LS

Stop PB

Transition C Fig. 20.5 State Transition Logic

Version 2 EE IIT, Kharagpur 11

State 1 Trans G St 3 St 2 Trans A St 5 St 6 Trans A State 1

St 4 Trans B

State 2 State 2 Trans B Trans D State 3 Trans C State 3 Trans C Fig. 20.6 State Logic This is followed by ladder logic to turn on outputs as requires by the steps. This section of ladder logic corresponds to the actions for each step. The rung for each output therefore contains one NO auxiliary contact corresponding to the state in which it is enabled. If an output is enabled at more than one state, the auxiliary NO contacts corresponding to those states would be connected in parallel. Similarly, if Manual switch or PB contacts are required, they also have to be put in parallel with the contacts for the states. Trans E State 4

State 2

Auto Indicator Lamp

Manual Sw. State 3 Dn Sol

Man Dn. Sol PB Fig. 20.7 Output Logic Some of the rungs corresponding to state, transition and output logics of the industrial stamping process are shown in Fig. 20.5-20.7. These are mostly self explanatory. The reader is advised to check the correctness of these rungs. Version 2 EE IIT, Kharagpur 12

Conclusion
Different methods can be used to design different controllers. The most basic controllers can be developed using simple flowcharts. More complex control problems should be solved with state diagrams. It is also possible to make a concurrent system using two or more state diagrams. However, for such concurrent processes the Sequential Function Chart formalism discussed in the next lesson is recommended.

Point to Ponder: 5
A. What would happen if the order of the transition, state and output logic blocks is changed in an RLL program? B. Mention any one disadvantage of a formal modelling approach, if you can think of it. C. Mention any one advantage of a formal modelling approach, apart from the reduced risk of programming errors.

Version 2 EE IIT, Kharagpur 13

Answers, Remarks and Hints to Points to Ponder Point to Ponder: 1
A. Before going through the rest of the system attempt to modify the RLL program shown in Fig. 20.2 to ensure that the process always moves to the up position first, and then from there, resumes its press cycle Ans: Without the state transition diagram this would be a considerable effort. However, with the state diagram approach of design discussed in this lesson, one only needs add a new state 7 between transition G and state 1 and then a new transition H between state 7 and state 1. This would imply that one new rung for state 7 and a new rung for transition H needs to be added. Transition G would now be a destination state for state 7 and not state 1. Similarly, Top LS would be the new enabling condition for transition H while the system would be in state 7. While in state 7, the UP solenoid would be on and therefore a new NO contact corresponding to state 7 needs to be added to the logic for UP solenoid in parallel with that for state 6. D. Before going through the rest of the system attempt to modify the RLL program shown in Fig. 20.2 to include a part detect sensor Ans: Already done in the lesson. E. Attempt to merge the above two solutions so that the new program is free from both the above problems Ans: Straightforward. Left as an exercise for the learner.

Point to Ponder: 2
A. Can we use corresponding switches, instead for the Auto, Stop and Reset PBs? Ans: The only difference between PBs and switches is that a PB is assumed to be released automatically, while the switch is not. However, since the transition logic does not remain on for more than one scan cycle, in this case, it does not make a difference whether these are PBs or switches. B. For the limit switches, and the part detect sensor, would you prefer mechanical switches over photo switches for this application? Justify. Ans: Mechanical switches would be preferred here, since these are much more rugged. Turning these on and off does not cause any problem for the ram which is hydraulically powered. C. Propose at least one more each of sensors, indicators and actuators for the above application and mention their benefits.

Version 2 EE IIT, Kharagpur 14

Ans: A hydraulic pressure sensor may be used for sensing the pump pressure before turning the machine into auto mode. This would be useful, particularly if the machine is run in auto mode for long intervals with robotic material handing equipment for part placement and removal. A fault indicator is also useful to indicate any deviation from normal cyclic operation. In this example, the electric power switch is always assumed to be on. One can add a contactor (actuator) to switch the power on, from state 1.

Point to Ponder: 3
A. Note that in step F above it is important to detect that the part is removed. What would happen, if this is not detected? Ans: The same part may be stamped many times, before it is removed. B. What would happen if after Stop PB is pressed, Reset PB and Auto PB are pressed in that sequence, even if the piston has not been taken to the top position manually? Ans: If reset PB is pressed the process would move to State 1. However, pressing the auto PB would not take it to State 2, since UP Limit switch would not be made. At this point the process would deadlock unless the piston is taken up manually. After it moves up, if the Auto PB is pressed, the machine would move to State 2.

Point to Ponder: 4
A. What would happen in the process controlled by the program shown in Fig. 20.2 if top/bottom Limit switch is stuck closed/open ? Ans: If the top LS is stuck closed, the Down solenoid would not be would off even if the piston comes to the bottom position. The piston would push the part to be stamped, but would not go up as desired during normal operation. Similarly, if the bottom LS is stuck closed, the piston would never come to the bottom position, since, the moment it would come to state 3, it would exit to state 4 and thus the UP solenoid would be on. One can similarly argue for the stuck open case. B. What would happen in the process controlled by the program shown in Fig. 20.2 if the hydraulic pump pressure becomes too low to move the RAM? Ans: The Up solenoid would be switched on, but the piston would not move. Thus the system would be deadlocked in state 4.

Point to Ponder: 5
A. What would happen if the order of the transition, state and output logic blocks is changed in an RLL program? Ans: If the transition block is not put first, there would be an unnecessary delay of one scan cycle in turning on the state. If the output block is put before the state logic there would be a similar delay. Version 2 EE IIT, Kharagpur 15

B. Mention any one disadvantage of a formal modelling approach, if you can think of it. Ans: The formal modelling may lead to more number of rungs in the RLL program compared to one that does not use it. Consequently the memory and execution time requirements would be higher. However, these are not significant drawbacks for present day PLC speed and memory sizes. C. Mention any one advantage of a formal modelling approach, apart from the reduced risk of programming errors. Ans: It is much easier to modify an RLL developed following the formal method.

Version 2 EE IIT, Kharagpur 16

Module 4
Programmable Logic Control Systems
Version 2 EE IIT, Kharagpur 1

Lesson 21
Programming of PLCs: Sequential Function Charts
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
After learning the lesson students should be able to A. Describe the major features of the IEC 1131-3 standard for PLC programming B. Describe the major syntax conventions of the SFC programming language C. Identify valid and invalid SFC segments D. Develop SFC programs for simple sequence control problems

Introduction
We have studied the RLL in a previous lesson. There are also other languages to program a PLC in, than the RLL. Most of the significant manufacturers support about 3 to 5 programming languages. Some of these languages, such as the RLL, have been in use for a long time. While most manufacturers used similar languages, these were not standardised in terms of syntactic features. Thus programs developed for one would not run in another without considerable modifications, often mostly syntactic. In the last few years there has been effort to standardise the PLC programming languages by the International Electrotechnical Commission (IEC). One of the languages, namely the Sequential Function Chart, which offers significant advantages towards development of complex structured PLC programs for concurrent industrial processes, is studied in detail. Know other languages are introduced in brief.

IEC 1131-3: The Language Standard

International

Programmable

Controller

IEC 1131 is an international standard for PLCs formulated by the International Electrotechnical Commission (IEC). As regards PLC programming, it specifies the syntax, semantics and graphics symbols for the following PLC programming languages:
• • • • •

Ladder diagram (LD) Sequential Function Charts (SFC) Function Block Diagram (FBD) Structured Text (ST) Instruction List (IL)

IEC 1131 was developed to address the industry demands for greater interoperability and standardisation among PLC hardware and software products and was completed in 1993. A component of the IEC 1131, the IEC 1131-3 define the standards for data types and programming. The goal for developing the standard was to propose a programming paradigm that would contain features to suit a large variety of control applications, which would eliminate proprietary barriers for the customer and their associated training costs. The language specification takes into account modern software engineering principles for developing clean, readable and modular code. One of the benefits of the standard is that it allows multiple languages to be used simultaneously, thus enabling the program developer to use the language best suited to each control task. Version 2 EE IIT, Kharagpur 3

Major Features of IEC 1131-3
The following are some of the major features of the standard. 1. Multiple Language Support: One of the main features of the standard is that it allows multiple languages to be used simultaneously, thus enabling the program developer to use the language best suited to each control task. 2. Code Reusability: The control algorithm can include reusable entities referred to as "program organization units (POUs)" which include Functions, Function Blocks, and Programs. These POUs are reusable within a program and can be stored in user-declared libraries for import into other control programs. 3. Library Support: The IEC-1131 Standard includes a library of pre-programmed functions and function blocks. An IEC compliant controller supports these as a "firmware" library, that is, the library is pre-coded in executable form into a prom or flash ram on the device. Additionally, manufacturers can supply libraries of their own functions. Users can also develop their own libraries, which can include calls to the IEC standard library and any applicable manufacturers' libraries. 4. Execution Models: The general construct of a control algorithm includes the use of "tasks", each of which can have one or more Program POUs. A task is an independently schedulable software entity and can be assigned a cyclic rate of execution, can be event driven, or be triggered by specific system functions, such as startup.

IEC 1131-3 Programming Languages
IEC 1131-3 defines two graphical programming languages (Ladder Diagram and Function Block Diagram), two textual languages (Instruction List and Structured Text), and a fifth language (Sequential Function Chart) that is a tool to define the program architecture and execution semantics. The set of languages include assembly-like low-level language like the Instruction List, as well as Structured Text having features similar to those of a high level programming language. Using these, different computational tasks of a control algorithm can be programmed in different languages, then linked into a single executable file. Below we first describe FBD, IL and ST in brief. The SFC is discussed in greater detail later in this lesson.

Function Block Diagram (FBD)
The function block diagram is a key product of the standard IEC 1131-3. FBD is a graphical language that lets users easily describe complex procedures by simply wiring together function blocks, much like drawing a circuit diagram with the help of a graphical editor. Function blocks are basically algorithms that can retain their internal state and compute their outputs using the persistent internal state and the input arguments. Thus, while a static mathematical function will always return the same output given the same input (e.g. sine, cosine), a function block can return a different value given the same input, depending on its internal state (e.g. filters, PID control). This graphical language clearly indicates the information or data flow among the different computational blocks and how the over all computation is decomposed among smaller blocks, each computing a well defined operation. It also provides good program documentation. The IEC 1131 standard includes a wide range of standard function blocks for performing a Version 2 EE IIT, Kharagpur 4

variety of operations, and both users and vendors can create their own. A typical simple function block is shown below in Fig. 21.1. Raise switch

&
Top LS OK IV Lower switch Loading valve

&
Bottom LS OK Pump running

&

=

Fig. 21.1 Combinational logic programmed with function blocks

Structured Text (ST)
Structured Text (ST) is a high-level structured programming language designed for expressing algorithms with complex statements not suitable for description in a graphical format. ST supports a set of data types to accommodate analog and digital values, times, dates, and other data. It has operators to allow logical branching (IF), multiple branching (CASE), and looping (FOR, WHILE…DO and REPEAT…UNTIL). Typically, a programmer would create his own algorithms as Functions or Function Blocks in Structured Text and use them as callable procedures in any program. A typical simple program segment written in Structured Text is shown below in Fig. 21.2. if (temp <= max_temp) then cool_valve :=false; m_vlv := (vlv23 +dbh18) /2; else alarm := true; end_if; Fig. 21.2 Simple program segment written in Structured Text.

Instruction List (IL)
A low-level assembly-like language, IL is useful for relatively simple applications, and works on simple digital data types such as boolean, integer. It is tedious and error prone to write large programs in such low level languages. However, because complete control of the implementation, including elementary arithmetic and logical operations, rests with the programmer, it is used for optimizing small parts of a program in terms of execution times and memory. A typical simple program segment written in Instruction List is shown below in Fig. 21.3. Version 2 EE IIT, Kharagpur 5

start_cmd: mul_op:

LD ii 01 ADD 10 MUL( i_gain SUB offset 01 ) ST op 01 JMPNC mul_op

Fig. 21.3 Simple program segment written in Instruction List

Point to Ponder: 1
A. Name one programming task for which, IL would be your chosen language. B. Name one programming task for which, ST would be your preferred language over FBD. C. For whom are the features code reusability and library support important, and why?

Sequential Function Chart (SFC)
SFC is a graphical method, which represents the functions of a sequential automated system as a sequence of steps and transitions. SFC may also be viewed as an organizational language for structuring a program into welldefined steps, which are similar conceptually to states, and conditioned transitions between steps to form a sequential control algorithm. While an SFC defines the architecture of the software modules and how they are to be executed, the other four languages are used to code the action logic that exercises the outputs, within the modules to be executed within each step. Similar modules are used for computation of the logical enabling conditions for each transition. Each step of SFC comprises actions that are executed, depending on whether the step is active or inactive. A step is active when the flow of control passes from one step to the next through a conditional transition that is enabled when the corresponding transition logic evaluates to true. If the transition condition is true, control passes from the current step, which becomes inactive, to the next step, which then becomes active. Each control function can, therefore, be represented by a group of steps and transitions in the form of a graph with steps labeling the nodes and transitions labeling the edges. This graph is called a Sequential Function Chart (SFC).

Steps
Each step is a control program module which may be programmed in RLL or any other language. Two types of steps may be used in a sequential function chart: initial and regular. They are represented graphically as shown below in Fig. 21.4.

Version 2 EE IIT, Kharagpur 6

(a)

(b)

Fig. 21.4 An initial step (a) and a regular step (b) in an SFC The initial step is executed the first time the SFC block is executed or as a result of a reset operation performed by a special function named SFC_RESET. There can be one and only one initial step in an SFC. The initial step cannot appear within a simultaneous branch construct, (which is described later in this section) but it may appear anywhere else. A regular step is executed if the transitional logic preceding the step makes the step active. There can be one or many regular steps in an SFC network, one or more of which may be active at a time. Only the active steps are evaluated during a scan. Each step may have action logic consisting, say, of zero or more rungs programmed in Relay Ladder Diagram (RLD) logic language. Action Logic is the logic associated with a step, i.e., the logic, programmed by RLL or any other logic, which is executed when the step is active. When a step becomes inactive, its state is initialised to its default state. A collection of steps may be labeled together as a macro-step.

Step

Action

(a)

(b)

Fig. 21.5 A step with action logic (a) and a macro-step (b) in an SFC

Transitions
Each transition is a program module like a step that finally evaluates a transition variable. Once a transition variable evaluates to true the step(s) following it are activated and those preceding it are deactivated. Only transitions following active states are considered active and evaluated during a scan. Transitions can also be a simpler entity such as a variable value whose value may be set by simple digital input. Transition logic can be programmed in any language. If programmed in RLL, each transition must contain a rung that ends with an output coil to set its transition variable.

Version 2 EE IIT, Kharagpur 7

S1 T1 S2 T2 S3 Fig. 21.6 Transitions connect steps in an SFC The SFC in Fig. 21.6 shows how the transitions connect steps in an SFC. Initially, step S1 is active. Thus transition T1 is also active. When the transition variable T1 becomes true, immediately, S1 becomes inactive, S2 becomes active while T1 becomes inactive and T2 becomes active.

Point to Ponder: 2
A. What is the difference between a step of an SFC and a state of an FSM? B. Why action logic is separately indicated from step logic, although both occur in the same step? C. How is the computation for step logic different from that of transition logic?

Basic Control Structures
Six basic control structures used in a sequential function chart are discussed below. .

Simple Sequence
In a simple sequence, control passes from step S2 to step S3 only if step S2 is active and transition T2 evaluates true.

S1 T1 S2 T2 S3 (a) Version 2 EE IIT, Kharagpur 8

Scan 1 2 3 4

S1 A I I I

T1 A I I I

S2 I A A I (b)

T2 I A A I

S3 I I I A

T3 I I I A

Fig. 21.7 A simple sequence in an SFC (a) and its execution over scans (b) The table in Fig. 21.7 (b) indicates the status (A : active; I:inactive) of te steps and transitions over scan cycles.

Divergence of a Selective Sequence
Divergence of a Selective Sequence means that after a step there is a choice of two or more transitions which can evaluate to be true. However, at a time, only one of which can be true, and therefore, the sequence of steps following that transition only are activated. For example, in the divergent selective sequence shown in Fig. 21.8, control passes from step S1 to step S2 if step S1 is active and transition T1 evaluates true. Control passes from step S1 to step S3 if step S1 is active, transition T1 is not true, and transition T2 is true (left-to-right priority of transition). Thus, at a time, exactly one branch of the selective sequence is selected. A left-to-right priority is used to determine the action branch if more than one transition evaluates true at the same time. A selective divergence must be preceded by one step. The first element after a selective divergence must be a transition. In terms of familiar programming constructs, this is similar to an IF…THEN…ELSEIF…ELSEIF….ELSE…construct.

S1

T1 S2 S3

T2

Fig. 21.8 Divergece of a selective sequence

Convergence of a Selective Sequence
A convergence of the divergent selective sequences must follow. Here transitions from each branch of the selective sequence converges eventually to one step. As shown in Fig. 21.9, in a convergent selective sequence, control passes from step S4 to step S6 if step S4 is active and transition T3 evaluates true. Similarly, control passes from step S5 to step S6 if step S5 is active and transition T4 is true.

Version 2 EE IIT, Kharagpur 9

S4 T3 S6

S5 T4

Fig. 21.9 Convergence of a selective sequence

Divergence of a Simultaneous Sequence
In contrast with a selective sequence, in a simultaneous divergent branch more than one step can become active. Thus, in this case, we have two or more branches to be active simultaneously. In the divergent simultaneous sequence shown in Fig. 21.10, control transfers from step S1 to step S2 and step S3 if step S1 is active and transition T1 evaluates true. Both steps S2 and S3 will become active. Note that the action logic of one step will be executed before the action logic of the other. The order of which step is executed first is undefined in the standard, and depends on a particular implementation. Note that the same thing happens for the state logic too, since computation is necessarily sequential in a single processor system. However the sequence becomes noticeable for action logic, since the output is observed in the field. A simultaneous divergent branch must be preceded by one transition. A step must be the first element after a simultaneous branch.

S1 T1

S2

S3

Fig. 21.10 Divergence of a simultaneous sequence

Convergence of a Simultaneous Sequence
A simultaneous convergent branch concludes a simultaneous sequence. It can only be preceded by step elements and not transitions as in the case of a selective sequence. It must be followed by one transition. In the convergent simultaneous sequence shown in Fig. 21.11, control passes from step S4 and step S5 to step S6 if steps S5 and S6 are both active and transition T2 evaluates true.

Version 2 EE IIT, Kharagpur 10

S4

S5

T2 S6 Fig. 21.11 Convergence of a simultaneous sequence The transition logic for T2 is only executed when all of the steps at the end of the simultaneous sequence are active.

Source and Destination Connectors
Source and destination connectors are used to create forward and backward jumps in an SFC. The keywords jump and cycle denote connectors in the SFC shown below. Backward jumps are called cycles. In the forward jump sequence shown in Fig. 21.12, control passes from step S2 to step S4 if step S2 is active, transition T2 evaluates to be false, and transition T3 evaluates to be true. In the backward jump (or cycle) sequence, control passes from step S4 to step S1 if step S4 is active and transition T5 evaluates true. Source and destination connectors cannot occur before a transition. Source connectors must occur immediately after the transition and indicate that the trasfer of control takes place once the transition fires. Destination connectors must occur immediately before a step indicating that the transfer of control after the transition before the corresponding source connector, this step becomes active.

Version 2 EE IIT, Kharagpur 11

cycle **

S1 T1 S2

T2 S3 T4 S4 T5 cycle *

T3 jump *

jump **

Fig. 21.12 Source and Destination Connectors Many PLCs also allow SFCs to entered be as graphic diagrams. Small segments of ladder logic can then be entered for each transition and action. Each segment of ladder logic is kept in a separate program. The architecture of such programs is discussed next.

Point to Ponder: 3
A. Identify whether the SFC segments indicated in Figs. 21.13-21.15 are valid. If not, justify your answer.

Version 2 EE IIT, Kharagpur 12

Cycle

S1 T1 S2 T2 S3

Cycle Fig. 21.13

S1 T1

S2 T2 S5

S3 T3 S6

S4 T4 S7

Cycle

T5 S8 T6 Cycle Fig. 21.14 Version 2 EE IIT, Kharagpur 13

S3 T2

S4

S5 T3

S6

S7

T4 S10

Fig. 21.15

Control Program Architecture with SFCs
A typical architecture of a control program with SFCs is shown in Fig. 21.16. Here the main program block is organised as an SFC. Each step and transition in the SFC of the main program block is coded as a module. These modules may be coded using any of the languages under the 1131-3 standard. These may be SFCs themselves. In general these modules may be organised in terms of a Preprocessing and a Post Processing block, in addition to the main sequential processing block.

Preprocessing
This section is processed at the start of every scan. Normally, RLD preprocessing logic is used to process, at the start of the scan cycle, events which may affect the sequential processing section of the program. These events may include: Initialization; Operator commands; Resetting the SFC to the initial state.

Version 2 EE IIT, Kharagpur 14

Sequential Processing
This portion of the PLC scan consists of evolving the SFC to its next state and processing the action logic of any steps that become active. Only the logic associated with active steps and transitions is scanned by the PLC, leading to a significant reduction of scan time.

Post processing
This section is processed every scan after the SFC is complete. It may contain Relay Ladder Diagram (RLD) logic to process safety interlocks, etc.
INPUTS

_MAIN PROGRAM BLOCK

SFC BLOCK 1

Preprocessing Sequential Processing Post processing Preprocessing Sequential Processing Post processing Preprocessing Sequential Processing Post processing

OUTPUTS

SFC BLOCK 2

SFC BLOCK 3

Fig. 21.16 Architecture of Control Software organized with SFCs Although SFCs are really needed for large and complex control problem, within the constraints of this lesson, we present below an SFC-based solution for the industrial stamping process control problem below.

Industrial Logic Control Example Revisited
Consider the industrial logic control example of the stamping process presented in Lesson 18 and further discussed in Lesson 20. Let us recall the description of the process given in the lesson. For convenience of reference the description and a pictorial representation of the process is reproduced below.

Version 2 EE IIT, Kharagpur 15

Piston

Up Solenoid Down Solenoid

Upper limit switch Die Lower limit switch

Fig. 21.17 An Industrial Logic Control Example

Summary of Requirements Analysis for the Stamping Process
A summary of the requirement analysis carried out in Lesson 20 is given below for ready reference.

Process Control Inputs
• • • • • • Part sensor: A position switch that detects when a part has been placed. Auto PB : A push button that indicates the ‘automatic mode’ Stop PB: A push button to stop the machine while the piston is moving down. Reset PB: After pressing Stop PB, this push button indicates that the machine is ready to return to stamping in the auto mode. Bottom LS: This sensor indicates the bottom position for the piston. Top LS: This sensor indicates top position for the piston.

Process Control Outputs
• • • • Up Solenoid: Drives the piston up. Down Solenoid: Drives the piston down. Auto Mode Indicator: Indicates ‘Auto’ mode. Part Hold: Holds the part during stamping.

Sequence of Events and Actions
A. The “Auto” PB turns the Auto Indicator Lamp on B. When a part is detected, the press ram advances down to the bottom limit switch C. The press then retracts up to top limit switch and stops Version 2 EE IIT, Kharagpur 16

D. A “Stop” PB, if pressed, stops the press only when it is going down E. If the “Stop” PB is pressed, the “Reset” PB must be pressed before the “Auto” PB can be pressed F. After retracting, the press waits till the part is removed and the next part is detected

State Transition Diagram
1 A 2 B 3 C D 5 Power/light off Dn Sol off 4 E 6 Dn Sol on Part hold Dn Sol off Up Sol on Up Sol off Part hold off F Part Removed Power

G

Fig. 21.18 The State Diagram for the industrial stamping press State No. 1 2 3 4 5 0 0 0 0 O/P Auto Indicator Part Hold Up Sol Down Sol 0 0 0 0 1 0 0 0 1 1 0 1 1 1 1 0 1 0 0 0 6

Fig. 21.19 The Output Table for the industrial stamping press

SFC-based Implementation of the Stamping Process Controller
The SFC for the industrial stamping process control problem is shown in Fig. 21.20. It is seen to be identical to the state diagram drawn for the process in Fig. 21.18, except possibly for graphical syntax used in the two diagrams. Thus for strictly sequential processes the SFC is nothing but the state diagram. However, it can also model concurrency, which cannot be captured in a simple FSM. There exist many DES modelling formalisms that capture concurrent FSM dynamics (such as Statecharts, widely used to model software dynamic modelling under Unified Modelling Language

Version 2 EE IIT, Kharagpur 17

1 6 7
Reset PB Auto PB

1

Part not detected

2
Part detect

Auto Indicator on

2
Down Sol on Part Hold on Bottom LS

3 4 5 3 4

Stop PB Auto Indicator off Down Sol off

Down Sol off Up Sol on Top LS

5

6

Up Sol off Part hold off

Figure 21.20 SFC for Controlling a Stamping Press Formalism). One can now develop the logic for the steps transitions and actions. In this lesson RLL is used for this purpose. Some of the ladder logic for the SFC is shown in Figure 21.x.

Version 2 EE IIT, Kharagpur 18

RLL for step 2 Auto Indicator

RLL for transition 2 Part detect EOT Step 2

RLL for step 3 Down Sol

Part hold RLL for transition 3 Bottom LS EOT Step 2

Fig. 21.21 Sample Ladder Logic for a Graphical SFC Program. Note the following distinctions of the SFC based implementation with that of one the state-based implementation with pure RLL. A. The initialisation logic can now be organised within the initial step of the SFC, which is explicitly meant for this purpose. Note that in the RLL implementation this had to be included within the logic for State 1. The SFC-based implementation is cleaner in this respect. B. The order of execution of the state and transition logic is determined by SFC execution semantics. Thus there is no need to worry about the order in which these program segments are physically ordered within the overall program. C. In the RLL implementation each and every step, transition and action logic is evaluated in every scan cycle. In SFC only the active states, their action logic and the active transitions are evaluated. This results in significant saving in processor time, which may be utilised in the system for other purposes. D. The ladder logic includes a new instruction, EOT, which will tell the PLC when a transition has completed. When the rung of ladder logic with the EOT output becomes true the SFC will move to the next step or transition.

Version 2 EE IIT, Kharagpur 19

Point to Ponder: 4
A. Develop an SFC for a two person assembly station. The station has two presses that may be used at the same time. Each press has a cycle button that will start the advance of the press. A bottom limit switch will stop the advance, and the cylinder must then be retracted until a top limit switch is hit. B. Create an SFC for traffic light control. The lights should have cross walk buttons for both directions of traffic lights. A normal light sequence for both directions will be green 16 seconds and yellow 4 seconds. If the cross walk button has been pushed, a walk light will be on for 10 seconds, and the green light will be extended to 24 seconds. C. Draw an SFC for a stamping press that can advance and retract when a cycle button is pushed, and then stop until the button is pushed again. D. Design a garage door controller using an SFC. The behavior of the garage door controller is as follows, a. there is a single button in the garage, and a single button remote control b. when the button is pushed the door will move up or down. c. if the button is pushed once while moving, the door will stop, a second push will start motion again in the opposite direction. d. there are top/bottom limit switches to stop the motion of the door.
e.

there is a light beam across the bottom of the door. If the beam is cut while the door is closing the door will stop and reverse.

f. there is a garage light that will be on for 5 minutes after the door opens or closes.

Version 2 EE IIT, Kharagpur 20

Answers, Remarks and Hints to Points to Ponder Point to Ponder: 1
A. Name one programming task for which, IL would be your chosen language. Ans: Consider a triple redundancy voting logic for 3 digital sensor inputs, for tolerance against sensor failures. The logic aims to select the three desired bits corresponding to the sensors from an input word. Then it evaluates a Boolean function that implements the voting logic (the exact Boolean logic for voting is left to the learner as an exercise). Note that all the above involve bit-operations on binary data types and are therefore easily and efficiently implemented using assembly like low-level languages. Hence the preference for IL. B. Name one programming task for which, ST would be your preferred language over FBD. Ans: Consider implementing a custom fuzzy-logic based PID controller. Since the controller involves logic and arithmetic based on real-valued data, RLLs are clearly not suited. Neither is IL, since the computations are complex and algorithmic in nature and based on real-valued data. Thus ST is the best suited for implementation of this algorithm C. For whom are code reusability and library support important, and why? Ans: These are very important for developers of control algorithms. This is because a proven library of routines of routines not only lead to faster development of control programs, they also lead to a better quality program in terms of a cleaner and more readable and reliable code.

Point to Ponder: 2
A. What is the difference between a step of an SFC and a state of an FSM? Ans: A step of an SFC denotes a computation module which gets executed cyclically, as long as the step is active. A state of an FSM is an instantiation of the values of its state variables. Note that a step in the SFC can represent a possibly cyclic subgraph of an FSM through which the FSM moves during the time the step is active. In the simplest case, a step of an SFC represents one state of an FSM. B. Why action logic is separately indicated from step logic, although both occur in the same step? Ans: The computation within a step can be of two types, namely, those that update internal state variables other than outputs and those that update the outputs which are exercised on physical outputs. The second type of computation is named action logic and explicitly indicated at the steps. Since the physical outputs are of final importance, these are separately indicated for each step. C. How is the computation for step logic different from that of transition logic? Version 2 EE IIT, Kharagpur 21

Ans: Systems are supposed to spend time in the states. Transitions are instantaneous and merely indicate conditions under which systems change from one state to another. In an SFC, at any point of time, a number of steps and transitions are active. However, the computations of the steps can update output variables in the action logic while the computations in the transitions cannot.

Point to Ponder: 3
A. Identify whether the SFC segments indicated in Figs. 21.13-21.15 are valid. If not, justify your answer. Ans: a. The SFC in Fig. 21.13 is invalid because a backward jump connects the two states S3 and S1 without an intervening transition. This is illegal. Any two steps in an SFC must contain an intermediate transition. b. The SFC in Fig. 21.14 is invalid because there is a jump into one of the branches of a simultaneous sequence. This is illegal, since the steps in the other branches of the simultaneous sequence are indeterminate. c. The SFC in Fig. 21.15 is valid.

Point to Ponder: 4
E. Develop an SFC for a two person assembly station. The station has two presses that may be used at the same time. Each press has a start cycle button that will start the advance of the press. A bottom limit switch will stop the advance, and the cylinder must then be retracted until a top limit switch is hit.

Version 2 EE IIT, Kharagpur 22

1. Master

0 Start button #1 1 press #1 adv. Bottom limit switch #1 2 press #1 retract Top limit switch #1 3 press #1 off

4 Start button #2 5 press #2 adv. Bottom limit switch #2 6 press #2 retract Top limit switch #2 7 press #2 off

F. Create an SFC for traffic light control. The lights should have cross walk buttons for both directions of traffic lights. A normal light sequence for both directions will be green 20 seconds and yellow 10 seconds. If the cross walk button has been pushed, a walk light will be on for 10 seconds, and the green light will be extended to 30 seconds.

Version 2 EE IIT, Kharagpur 23

2. Start

EW crosswalk button Red NS, green EW walk light on for 10s 30s delay

NO EW crosswalk button Red EW, green NS 20s delay

Red EW, yellow NS

10s delay NS crosswalk button Red EW, green NS walk light on for 10s 30s delay NO NS crosswalk button Red NS, green EW 16s delay

Red NS, yellow EW

4s delay

G. Draw an SFC for a stamping press that can extend and retract when a cycle button is pushed, and then stop until the button is pushed again.

Version 2 EE IIT, Kharagpur 24

3. Start

idle Cycle button pressed extending Advance limit switch made retracting Retract limit switch made

H. Design a garage door controller using an SFC. The behavior of the garage door controller is as follow. a. There is one button inside the garage, and one button remote control. b. When either of the buttons are pushed the door will move up or down. c. If the button is pressed once while moving, the door will stop, a second press will start motion again in the opposite direction. d. There are top/bottom limit switches to stop the motion of the door when it reaches either of the two ends.
e. f.

There is an infrared beam across the bottom of the door. If the beam is interrupted while the door is closing the door will stop and reverse. There is a garage light that will be on for 5 minutes after the door opens or closes.

Version 2 EE IIT, Kharagpur 25

4. 1

2 T1 Garage or Remote button pressed 3 Close door

T3 Light beam interrupted

T2

bottom limit switch made

4 T4 Garage or Remote button pressed

5 T5

Open door

top limit switch made

Version 2 EE IIT, Kharagpur 26

Module 4
Programmable Logic Control Systems
Version 2 EE IIT, Kharagpur 1

Lesson 22
The PLC Hardware Environment
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
After learning the lesson students should be able to A. Describe the physical organization of hardware in the PLC B. State typical components and functionality of the main types of modules C. Describe typical Function modules used in PLC systems

Introduction
In Lesson 18 the architecture of a PLC system has been presented. In this lesson the hardware characteristics of the components of a PLC system and their physical organization are discussed in some detail. PLC systems are available in many hardware configurations, even from a single vendor, to cater to a variety of customer requirements and affordability. However, there are some common components present in each of these. These components are: A. Power Supply - This module can be built into the PLC processor module or be an external unit. Common voltage levels required by the PLC are 5Vdc, 24Vdc, 220Vac. The voltage lends are stabilized and often the PS monitors its own health. B. Processor - This is the main computing module where ladder logic ands other application programs are stored and processed. C. Input/Output - A number of input/output modules must be provided so that the PLC can monitor the process and initiate control actions as specified in the application control programs. Depending on the size of the PLC systems the input-output subsystem can either span across several cards or even be integrated on the processor module. Some of there input-output Input/output cards generates/accept TTL level, clean signals. Output ‘modules’ provide necessary power to the signals. Input ‘modules’ converts voltage levels, cleans up RF noise and isolates it from common mode voltages. I/O modules may also prevent over voltages to reach the CPU or low level TTL. D. Indicator lights - These indicate the status of the PLC including power on, program running, and a fault. These are essential when diagnosing problems. E. Rack, Slot, Backplane – These physically house and connect the electronic components of a PLC.

In addition there may be
A. Communication processors that realize remote and network i/o for PLCs B. Programming and man-machine interface devices for PLCs.

Version 2 EE IIT, Kharagpur 3

Figure 22.1 shows the typical subsystems found on a PLC system.
MMI Memory Analog I/O Modules Function Modules Communication Processor Remote I/O Modules Digital I/O Modules

End User

CPU

Back Plane Power Supply

Programmer Unit

Printer

Fig. 22.1 Typical Subsystems for a PLC system

Point to Ponder: 1
A. What is remote i/o? How is it different from the other kinds of i/o? B. What are the functions of the blocks named MMI and Programmer? The configuration of the PLC refers to the physical organization of the components. Typical configurations are listed below from largest to smallest. A. Rack - A rack is often large and can hold multiple cards. These cards, which realize the CPU, power, communication, i/o and special function modules, are connected by a bus, often called a backplane. When necessary, multiple racks can be connected together by bus extenders. Each channel in a card can be addressed by a rack – slot – channel addressing scheme, which varies from vendor to vendor. These tend to be of highest cost, but also the most flexible and easy to maintain. The functional architecture of such a rack mounted PLC system is shown in Fig. 22.2. The figure shows the various types of functional subsystems, which may or may not be on the same board, connected through a backplane. However this does not reflect the physical organization the various modules that make a PLC system. This is shown, distributed over a number of racks along with bus extension system shown in Fig. 22.3. The figure shows that while direct connection may be possible for extension over small distances of a few meters. For extension over longer distances special bus extension units are needed to provide the necessary drives for reliable signal transmission over a distance.

Version 2 EE IIT, Kharagpur 4

Memory Modules

Central Processor

Co Processor

IO Processor

Data Bus
Communication Processor Intelligent IO Module Input Module Output Module

Fig. 22.2 Functional hardware organization of a PLC System
Central Expansion Unit Expansion Unit Central Controller max. 2m max. 200 m max. 1 km Expansion Unit max. 2m Expansion Unit Expansion Unit Remote Interfacing Decentral Expansion Unit Expansion Unit Expansion Unit

Fig. 22.3 Physical layout and Bus extension System for PLCs B. Mini - These are similar in function to PLC racks, but about half the size. Photograph of one such PLC system is shown in Fig. 22.4. These generally are situated completely at one place and do not use an extended bus. Floor mounted or wall mounted.

Fig. 22.4 A mini PLC System Version 2 EE IIT, Kharagpur 5

C. Compact - A compact, all-in-one unit (about the size of a shoebox) that has limited expansion capabilities. Lower cost, and compactness make these ideal for small applications. Usually wall mounts. D. Micro - These units can be as small as a deck of cards suitable for wall mounted or table top. They tend to have fixed quantities of I/O and limited abilities, but costs will be the lowest. Used for simple embedded applications. Often not suitable for industrial applications. E. Software - A software based PLC requires a general purpose computer, like a PC, with an interface card. The software, utilizes the operating system resources of the computer to realize control, logic and i/o functions. An advantage of such a configuration is that it allows the PLC to be connected to sensors, other similar PLCs or to other computers across a general purpose network, such as the ethernet. The PLC can also function concurrently with other PC-based applications like a visualization software.

Point to Ponder: 2
A. Name one application each for which a mini PLC may be appropriate. Provide justifications for your choice. B. Why is a special bus extender unit needed for extending the bus over long distances?

Processor Module
A wide range of processor modules, scalable in terms of performance and capacity, are available to meet the different needs of users. Processors manage the whole PLC station consisting of discrete input/output modules, analog modules and application-specific function modules (counting, axis control, stepper control, communication, etc.) located on one or more racks connected to the backplane. In terms of hardware, besides a CPU and possible co-processor, each processor module typically includes: • • • • • a protected internal RAM memory which can take the application program and can be extended by memory extension cards (RAM or Flash EPROM) a realtime clock ports for connecting several devices simultaneously for purposes such as programming, human-machine interface etc. communication cards for various industrial communication standards such as, Modbus+ or Fieldbus, as well as serial links and Ethernet links Display block with LEDs, RESET button, used to activate a cold restart of the PLC system.

Typical specifications for a high end and a low end PLC processor module for a rack-based PLC system are given below. Features No. of racks No. of module slots In-rack discrete I/O Low end 6 21 512 High end 24 87 2048

Version 2 EE IIT, Kharagpur 6

In-rack analog I/O

Application specific modules Process control loops Process control channels Network connection: TCP/IP, Modbus +, Ethernet Fieldbus connection Internal memory (16-bit words) Memory extension (16-bit words)

24 function 8 1 0 32K 64K

256 64 60 20 4 2 176K 512K

Table 22.1 Typical Features of high end and low-end processor modules Processor modules contain function block libraries, which can be configured to work with other modules, to realize various automation related functionality, such as, • • • Counting up to 10 – 100 KHz PID Control with algorithms realized in different forms Controlled positioning for manufacturing by CNC machines with stepper / servo drives, and features such as rapid traverse / creep speed for high accuracy positioning of point to point axes, interpolation and multi axis synchronization for contouring axes Input/output: These may be categorized as digital / analog depending on the nature of the signal or as local/remote/networked, depending on the interface through which it is acquired. These are described in detail below.

Input Module
Input modules convert process level signals from sensors (e.g. voltage face Contacts, 0-24v Dc, 4 – 20mA), to processor level digital signals such as 5V or 3.3 V. They also accept direct inputs from thermocouples and RTDs in the analog case, and limit switches or encoders in the digital case. Naturally, therefore these modules include circuitry for galvanic isolation, such as those using optocouplers. Galvanic isolation-

Analog input modules
Analog input modules convert analog process level signals to digital values, which are then processed by the digital electronic hardware of the programmable controller. A set of typical parameters that define an analog input module are shown in Table 22.2. The analog modules sense 8/16 analog signals in the range ± 5 V, ± 10 V or 0 to 10 V. Each channel can either be single-ended, or differential. For single ended channels only one wire is connected to a channel terminal. The analog voltage on each channel terminal that is sensed is referred to a common ground. In the case of differential channels, each channel terminal involves two wires and the voltage between the pair of wires is sensed. Thus both the wires can be at different voltages and only their difference is sensed and converted to digital. Differential channels are more accurate but consume more electronic resources of the module for their processing. Often these modules Version 2 EE IIT, Kharagpur 7

also house channels that output analog/digital signals, as well as excitation circuitry for sensors such as RTDs. An analog module typically contains: Analog to digital (A/D) converters Analog multiplexers and simultaneous sample-hold (S/H) Analog Signal termination PLC bus ports Synchronisation

Fig. 22.5 Analog IO Module Module Parameter Number of input Galvanic isolation Input ranges Input impedance for various ranges (ohm) Types of sensor connections Data format Conversion principle Conversion time Type/Number/Typical Value 8/16 voltage/current/Pt 100/ RTD Yes /No ±50 mV to ±10 V; ±20 mA; Pt 100 ±50mV: > 10 M ; ±10 V: > 50k; ±20 mA : 25; Pt 100 : > 10 M 2-wire connection; 4-wire connection for Pt 100 11 bits plus sign or 12 bit 2’s complement Integrating /successive approximation In ms (integrating) , μs (successive approx.)

Table 22.2 Typical Parameters for an Analog Input Module

Digital Input Modules
The digital inputs modules convert the external binary signals from the process to the internal digital signal level of programmable controllers. Digital input channel processing involves isolation and signal conditioning before inputting to a comparator for conversion to a 0 or a 1. The typical parameters that define a digital input module are shown in tabular form along with typical values in Table. Version 2 EE IIT, Kharagpur 8

Module Parameter Number of input Galvanic isolation Nominal input voltage Input voltage range - “0” signal - “1” signal Input current Delay Maximum cable length

Typical Values 16/32 yes + 24 V DC -33…+7V +13…+33V Typically in mA Typically in μs Typically within 1000m

Table 22.3 Typical Parameters for an Digital Input Module

Point to Ponder: 3
A. Why so many types of analog inputs have been provided for? B. How to select between integrating and successive approximation converters? C. What is the significance of the input ranges for digital input modules?

Output Modules
Outputs to actuators allow a PLC to cause something to happen in a process. Common actuators include: 1. Solenoid Valves - logical outputs that can switch a hydraulic or pneumatic flow. 2. Lights - logical outputs that can often be powered directly from PLC boards. 3. Motor Starters - motors often draw a large amount of current when started, so they require motor starters, which are basically large relays. 4. Servo Motors - a continuous output from the PLC can command a variable speed or position to a servo motor drive system. The outputs from these modules may be used to drive such actuators. Consequently, they include circuitry for current / power drive using solid-state electronics such as transistors for DC outputs or triacs for AC outputs. Continuous outputs require output cards with D/A converters. Sometimes they also provide potential free relay contacts (NO/NC), which may be used to drive higher power actuators using a separate power source. Since these modules straddle across the processor and the output power circuit, these must provide isolation. However, most often, output modules act as modulators of the actuator power, which is actually applied to the equipment, machine or plant. External power supplies are connected to the output card and the card will switch the power on or off for each output. Typical output voltages are 120V ac, 24V dc, 12-48V ac/dc, 5V dc (TTL) or 230V ac. These cards typically have 8 to 16 outputs of the same type and can be purchased with different current ratings. A common choice when purchasing output cards is relays, transistors or triacs. Relays are the most flexible output devices. They are capable of switching both AC and DC outputs. But, they are slower (about Version 2 EE IIT, Kharagpur 9

10ms switching is typical), they are bulkier, they cost more, and they wear out after a large number of cycles. Relays can switch high DC and AC voltage levels while maintaining isolation. Transistors are limited to DC outputs, and triacs are limited to AC outputs. Transistor and triac outputs are called switched outputs. In this case, a voltage is supplied to the PLC card, and the card switches it to different outputs using solid-state circuitry (transistors, triacs, etc.). Triacs are well suited to AC devices requiring less than 1A. Transistor outputs use NPN or PNP transistors up to 1A typically. Their response time is well under 1ms.

Analog Output Module
Analog output modules convert digital values from the PLC processor module into an analog signal required by the process. These modules therefore require a D/A converter for providing analog outputs. However, typically, servo-amplifiers for power amplification, required for driving high current loads directly, are not integrated on-board. Front connectors are used for terminating the signal cables. Modules and front connectors may be inserted and removed under power. The output signals can be disabled by means of an enable input. The last value then remains latched. Typical parameters that define an analog output module are shown in Table 22.4 along with typical values.

Digital Output Module
Digital output modules convert internal signal levels of the programmable controllers into the binary signal levels required externally by the process. Output can be DC or AC. Up to 16 outputs can be connected in parallel. Indication for short-circuits, fuse blowing etc. are often provided. Number of outputs Galvanic isolation Output ranges ( rated values ) Load resistance - for voltage outputs min. - for current outputs max. Digital representation of the signal Conversion time Short-circuit protection Short-circuit current approx. Open-circuit voltage approx. Linearity in the rated range Cable length max. 8 voltage and current output yes ± 10 V; 0…20 mA 3.3 k 300 11 bits plus sign In μs yes 25 mA (for a voltage output) 18 V (for a current output) ±0.25% + 2 LSB 200 m

Table 22.4 Typical Parameters for an Analog Output Module The typical parameters that define a digital output module are shown in Table 22.5 along with typical values.

Version 2 EE IIT, Kharagpur 10

Module Parameter 1.Number of outputs 2.Galvanic isolation Rated value of Supply voltage Permissible range Max. output current for “1” signal Short-circuit protection Max. switching frequency for resistive loads, lamps, inductive loads, respectively, in Hz. “0” signal level max “1” signal level max. Max. cable length (unshielded)

Typical Value 16/32 yes + 24 V DC 20-30 V 0.5 A Yes 100/11/2 Hz ( at 0.3 A ) +3V Vpp-1.5 V 400 m

Table 22.5 Typical Parameters for an Digital Output Module

Point to Ponder: 4
A. Determine the significance of the following specifications for an analog output module: a. Load resistance b. Linearity c. Conversion time B. Determine the significance of the following specifications for a digital output module: a. Max. switching frequency b. Shrot circuit protection

Function Modules
For high speed i/o tasks such as one required to measure speed by counting pulses from shaft angle encoders, or for precision position control applications, independent i/o modules that execute tasks independently of the central processor are required to meet the timing requirements of the i/o. The signal preprocessing “intelligent” I/O- modules make it possible to count fast impulse trains, to acquire and process travel increments, speed and time measure etc., i.e. they take on the critical timing control tasks which normally can’t be carried out fast enough by the central processor with its programmable logic control, as well as its primary logic control functions. These modules not only relieve the central processor of additional tasks, they also provide fast and specialized solutions to some common control problems. The processing of the signals is carried out primarily by the appropriate I/O- modules, which frequently operate with their own processor. Below we discuss two such modules, which are used with PLCs to handle specific high performance automation functions. A. A Count Module is employed where pulses at high frequency have to be counted, i.e. when machines run fast. It can also be applied to output fast pulse trains or realize accurate timing signals. B. A Loop Controller Module is primarily used where high speed closed loop control is required, such as with controlled drives. The preprogrammed, parameterized functions available with the module (e.g. for ramp-function generation, speed regulation, signal limit monitoring) can be easily parameterized via graphical interfaces by a programmer. Version 2 EE IIT, Kharagpur 11

Count Module
A count module senses fast pulses, from sources such as shaft angle encoders, through several input ports. Counting frequency can be as high as 2 MHz and a typically, a counter of length 16 bit or more can count up and down. Counter modules can often also be applied for time and frequency measurement and as a frequency divider.

Fig. 22.6 A high speed counter module Typical counter module hardware contains, among possible other things, an interface to the processor through the system bus, a counter electronics block, a quartz controlled frequency generator and a frequency divider. For example, it may contain, say, 5 counters with, say, 16 bits, each of which are cascadable. In this way, up to 80 bit can be counted in various codes. Thus, decimals up to about 1024 can be counted. Each port input can be switched on to the counter at random. It is possible to place a frequency divider from 1 to 16, between the port input and the counter. The frequency of an internal frequency generator can be directed either straight to a counter or via the frequency divider to a port input. On reaching the terminal count, the counter outputs a level or edge signal. For each counter there are a number of different operating modes, which can be set by a user program. With a comparator and an alarm register, a number of count values can be compared and under defined conditions configured to turn on a process alarm. A counter can be programmed in many ways, such as: Count mode binary or BCD coded Count once or cyclically Count on rising or falling edge Count up or down Counting of internal clock or external pulses

Version 2 EE IIT, Kharagpur 12

Loop Controller Module
A loop controller module is suitable for solving fast control loop problems. A typical module can process several control loops with sampling times varying between a few milliseconds to several seconds. The process output values are measured via analog input ports and are compared with the set point values. The power circuits of the actuator units are driven through analog output ports. Such a module contains a microprocessor, which controls the sensing and processing of the process output and set point values and computes the control law and outputs the manipulated variables. The operating configurations of the loop controller module are assigned with a programmer and committed to memory located on the module. The central controller provides set point values, parameters and control commands and reads the output values. The application software for the module can be structured in terms of standard close loop functions (e.g. ramp function generator, speed controller, etc.). These standard functions can be interconnected to a closed loop structures with the aid of a programmer interface and are the resulting control code automatically compiled and downloaded to memory on the module. The microprocessor executes the standard functions in accordance with the designed closed loop structure for an application such as a motor drive or a standard cascade process control loop. A drive loop controller would comprise all necessary functions for the control of a drive, while a cascade loop controller would consist of a cascade of two control loops. The outer loop controller can, for an instant, be used for closed loop position control, while the inner loop for control of rotational speed control. Each loop controller can be equipped with P, D, PI, PD or PID algorithms. As additional functions there may exist limit monitoring indicators limit monitoring indicators, which for example, can monitor the actual armature current value, for thermal supervision of the drive.

Point to Ponder: 5
A. Name two advantages and two disadvantages of using a function module for an automation application B. Based on the illustrative figures given for the counter module, determine the maximum possible timing interval that can be programmed on the counter module. C. Describe the meaning of the various counting modes of a counter module D. Consider a position control application. Describe the options for receiving feedbacks for the loop controller for this application

Lesson Summary
In this lesson, the following topics related to PLC System hardware have been discussed. A. Introduction to typical PLC hardware subsystems B. Physical organization of the PLC hardware subsystems C. Typical features of Processors Modules D. Typical features of analog and digital input modules E. Typical features of analog and digital output modules F. Two function modules : High Speed Counter and Loop Controller Version 2 EE IIT, Kharagpur 13

Answers, Remarks and Hints to Points to Ponder Point to Ponder: 1
A. What is remote i/o? How is it different from the other kinds of i/o? Ans: Local i/o, as contrasted with remote i/o is where the field terminals of the PLC i/o modules are connected directly to the field devices. Each channel carries data that is not multiplexed. On the other hand for remote i/o, multiplexed data for several field channels is sent in multiplexed form to a remote i/o device that demultiplexes and transmits data for each field device to it, with or without data conversion and signal conditioning. Local i/o may be analog or digital. Remote i/o is always digital. Remote i/o is used to mainly to save on cabling of individual data channels from the device to the PLC rack. B. What are the functions of the blocks named MMI and Programmer? Ans: MMI is an acronym for Man Machine Interface. Generally PLC modules do not have faciities for visualization. However, if needed, one can connect a special MMI device like a terminal or a printer and visualization of process outputs and their transitions. Similarly, a programmer is another device which facilitates easy development, debugging and monitoring of programs through graphical interfaces. The developed programs can be compiled and downloaded into the PLC memory. Programmers are available in table-top PC based as well as handheld versions.

Point to Ponder: 2
A. Name one application each for which a mini PLC may be appropriate. Provide justifications for your choice. Ans: A typical example would be a CNC machining Centre. Typical i/o requirements for such a centre would be less than 100 channels. The number of channels are going to be more or less fixed. Communication, visualisation and programming requirements are non trivial but not extensive either. Therefore, neither a full PLC rack, nor a compact PLC is suitable. B. Why is a special bus extender unit is needed for extending the bus over long distances? Ans: That is because for long distances, the capacitive loading of connecting cables increases significantly. A much higher value of capacitance therefore needs to charged and discharged at a high rate, without causing signal voltage degradations. This requires much higher current driving capability and thus necessitates a separate driver module.

Point to Ponder: 3
A. Why so many types of analog inputs have been provided for?

Version 2 EE IIT, Kharagpur 14

Ans: So that the user need not face the trouble of signal conditioning for most of the common sensing devices, such as those providing voltage or current outputs, resistance sensors, thermocouples. These can be easily interfaced with the input modules directly. B. How to select between integrating and successive approximation converters Ans: For integrating ADCs conversion is the slowest with conversion times in the range of several milliseconds but can be very accurate. Successive approximation types are a thousand times faster, are now available with good accuracies and therefore are suited to most applications and are popular too. C. What is the significance of the input ranges for digital input modules? Ans: The wide input voltage ranges and the wide separation of the ranges for 0 and 1 signal levels has been designed to ensure reliable data transmissin in possibly very noisy industrial environments. Obviously power levels for signals drivers have to be very high to maintain these levels but is not a concern here.

Point to Ponder: 4
A. Determine the significance of the following specifications for an analog output module: a. Load resistance b. Linearity c. Conversion time Ans: The significance of the specifications are described below. • • • Load Resistance: Decides the current drive capacity that needs to be provided on the output module Linearity: Basically works like an accuracy specification between the digital value and the analog field level output. Conversion time: Added to the control computation time, decides the lower bound on sampling time that can be used for control.

B. Determine the significance of the following specifications for a digital output module: a. Max. switching frequency b. Shrot circuit protection • Maximum switching frequency: Decides the speed of the control application. However, for inductive loads, the switching frequency depends on the power sourcing and sinking capabilities provided on the module. Short circuit protection: Since the load being driven may be shorted, this feature is essential to protect the module from damage.

Point to Ponder: 5
A. Name two advantages and two disadvantages of using a function module for an automation application

Version 2 EE IIT, Kharagpur 15

Ans: First advantage is that the main processor computational burden is reduced significantly. For example, for a control function, that main processor only has to supply the set point and not compute the control input every sampling time. The second advantage is that the application program developer does not need to be concerned with the fine details of a control algorithm for a specific application such as a precision positioning. One of the disadvantages is the additional cost of such a module. The second disadvantage may be that the functional module may prove to be a constraint if one is interested in using a different control algorithm for the application. E. Based on the illustrative figures given for the counter module, determine the maximum possible timing interval that can be programmed on the counter module. Ans: The maximum timing interval that can be programmed on counter, for the data of 5 cascadable 16 bit counters, is, the half clock period divided by 280. F. Describe the meaning of the various counting modes of a counter module Ans: The meanings of the modes are described below. Count mode binary or BCD coded: Refers to the code using which the count values propagate. Codes may be chosen depending on requirements of other software/hardware modules that interface to it. Count once or cyclically: Refers to whether the counter should stop after reaching terminal count or would reload the initial count and restart counting. Count on rising or falling edge: Refers to the edge that increments/decrements count value. Count up or down : Self explanatory Counting of internal clock or external pulses: If the counter increments by internal clock, it becomes a timer. It would generate output after a fixed time interval which depends on the clock frequency and the number of bits in the counter. With external pulses it counts the number of pulse events in the field. G. Consider a position control application. Describe the options for receiving feedbacks for the loop controller for this application Ans: The options for feedback depend on the sensors used in the drive. If the position sensor used is analog, like a resolver, a potentiometer or an LVDT, then an analog input channel has to be used. On the other hand, if the sensor used is a shaft angle encoder, then either a digital input channel is to be used, in which case, the pulse counting has to be done by the module. Else the encoder channel can be interfaced to a counter function module and the count value read from it via the bus. Similar arguments apply for the speed feedback input, if any.

Version 2 EE IIT, Kharagpur 16

Module 5
CNC Machines
Version 2 EE IIT, Kharagpur 1

Lesson 23
Introduction to Computer Numerically Controlled (CNC) Machines
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
After learning the lesson students should be able to A. Define Numerical Control and describe its advantages and disadvantages B. Name and describe the major components of a CNC system C. Explain the coordinate systems adopted for CNC programming D. Describe the major types of motion control strategies E. Describe the major classifications of CNC machines

Introduction

Cutter Spindle Workpiece Spindle

Tool

Workpiece

Table

(a) Turning

(b) Milling

Fig. 23.1 Drive in a metal cutting

Introductory Concepts of Machining
Machining is basically removal of material, most often metal, from the workpiece, using one or more cutting tools to achieve the desired dimensions. There are different machining processes, such as, turning, milling, boring etc. In all these cases metal is removed by a shearing process, which occurs due to the relative motion between the workpiece and the tool. Generally, one of the two rotates at designated and generally high speed, causing the shearing of material (known as chips), from the workpiece. The other moves relatively slowly to effect removal of metal throughout the workpiece. For example, as seen above in a turning operation of lathes, the “job” or the workpiece rotates in a chuck, while the tool moves in two dimensions translationally. On the other hand, in milling, it is the cutter which rotates on a spindle, while the workpiece, which is fastened to a table, moves in X-Y dimensions. While, a precise and high speed rotational motion is needed for good finish of the machined surface, for dimensional accuracy, precise position and velocity control of the table drive are essential.

Version 2 EE IIT, Kharagpur 3

TOOL MOTION

WORKPIECE MOTION

Shaping broaching Broaching Sawing planing Turning boring Hobbing Milling grinding

Drilling boring

Stationary or intermittent motion Rectilinear motion

Rotary motion

Resultant of rotary and rectilinear motion Fig. 23.2 Nature of motion of the Job and the Tool for various Metal Cutting Processes For all metal-cutting processes, the cutting speed, feed, and depth of cut are important parameters. The figure below shows the important geometry for the turning process. The cutting speed, which is a measure of the part cut surface speed relative to the tool. Speed is a velocity unit for the translational motion, which is may be stated in or meters/min. The depth of cut, DOC is the depth that the tool is plunged into the surface. Feed defines the relative lateral movement between the cutting tool and the workpiece. Thus, together with depth of cut, feed decides the cross section of the material removed for every rotation of the job or the tool, as the case may be. Feed is the amount of material removed for each revolution or per pass of the tool over the workpiece and is measured in units of length/revolution, length/pass or other appropriate unit for the particular process.

Version 2 EE IIT, Kharagpur 4

Speed

depth of cut Feed

Fig. 23.3 Speed, feed and depth of cut for turning operation

Point to Ponder: 1
A. Consider the metal cutting configuration for milling shown. Identify the motion directions corresponding to speed, feed and depth of cut

B. “Metal cutting is essentially a shearing phenomenon”, justify the statement with respect to the entries in Table 23.1.

What is Computer Numerical Control?
Modern precision manufacturing demands extreme dimensional accuracy and surface finish. Such performance is very difficult to achieve manually, if not impossible, even with expert operators. In cases where it is possible, it takes much higher time due to the need for frequent dimensional measurement to prevent overcutting. It is thus obvious that automated motion control would replace manual “handwheel” control in modern manufacturing. Development of computer numerically controlled (CNC) machines has also made possible the automation of the machining processes with flexibility to handle production of small to medium batch of parts. In the 1940s when the U.S. Air Force perceived the need to manufacture complex parts for highspeed aircraft. This led to the development of computer-based automatic machine tool controls also known as the Numerical Control (NC) systems. Commercial production of NC machine tools started around the fifties and sixties around the world. Note that at this time the microprocessor has not yet been invented. Version 2 EE IIT, Kharagpur 5

Initially, the CNC technology was applied on lathes, milling machines, etc. which could perform a single type of metal cutting operation. Later, attempt was made to handle a variety of workpieces that may require several different types machining operations and to finish them in a single set-up. Thus CNC machining Centres capable of performing multiple operations were developed. To start with, CNC machining centres were developed for machining prismatic components combining operations like milling, drilling, boring and tapping. Gradually machines for manufacturing cylindrical components, called turning centers were developed.

Numerical Control
Automatically controlling a machine tool based on a set of pre-programmed machining and movement instructions is known as numerical control, or NC. In a typical NC system the motion and machining instructions and the related numerical data, together called a part program, used to be written on a punched tape. The part program is arranged in the form of blocks of information, each related to a particular operation in a sequence of operations needed for producing a mechanical component. The punched tape used to be read one block at a time. Each block contained, in a particular syntax, information needed for processing a particular machining instruction such as, the segment length, its cutting speed, feed, etc. These pieces of information were related to the final dimensions of the workpiece (length, width, and radii of circles) and the contour forms (linear, circular, or other) as per the drawing. Based on these dimensions, motion commands were given separately for each axis of motion. Other instructions and related machining parameters, such as cutting speed, feed rate, as well as auxiliary functions related to coolant flow, spindle speed, part clamping, are also provided in part programs depending on manufacturing specifications such as tolerance and surface finish. Punched tapes are mostly obsolete now, being replaced by magnetic disks and optical disks. NC equipment has been defined by the Electronic Industries Association (E1A) as “A system in which actions are controlled by the direct insertion of numerical data at some point. The system must automatically interpret at least a portion of this data.” This is an old definition as is apparent from the terminology used in the definition Computer Numerically Controlled (CNC) machine tools, the modern versions of NC machines have an embedded system involving several microprocessors and related electronics as the Machine Control Unit (MCU). Initially, these were developed in the seventies in the US and Japan. However, they became much more popular in Japan than in the US. In CNC systems multiple microprocessors and programmable logic controllers work in parallel for simultaneous servo position and velocity control of several axes of a machine for contour cutting as well as monitoring of the cutting process and the machine tool. Thus, milling and boring machines can be fused into versatile machining centers. Similarly, turning centers can realize a fusion of various types of lathes. Over a period of time, several additional features were introduced, leading to increased machine utilisation and reduced operator intervention. Some of these are: (a) Tool/work monitoring: For enhanced quality, avoidance of breakdowns. (b) Automated tool magazine and palette management: For increased versatility and reduced operator intervention over long hours of operation (c) Direct numerical control (DNC): Uses a computer interface to upload and download part programs in to the machine automatically.

Version 2 EE IIT, Kharagpur 6

Advantages of a CNC Machine
CNC machines offer the following advantages in manufacturing. • • Higher flexibility: This is essentially because of programmability, programmed control and facilities for multiple operations in one machining centre, Increased productivity: Due to low cycle time achieved through higher material removal rates and low set up times achieved by faster tool positioning, changing, automated material handling etc. Improved quality: Due to accurate part dimensions and excellent surface finish that can be achieved due to precision motion control and improved thermal control by automatic control of coolant flow. Reduced scrap rate: Use of Part programs that are developed using optimization procedures Reliable and Safe operation: Advanced engineering practices for design and manufacturing, automated monitoring, improved maintenance and low human interaction Smaller footprint: Due to the fact that several machines are fused into one.

• • •

On the other hand, the main disadvantages of NC systems are • • • Relatively higher cost compared to manual versions More complicated maintenance due to the complex nature of the technologies Need for skilled part programmers.

The above disadvantages indicate that CNC machines can be gainfully deployed only when the required product quality and average volume of production demand it.

Classification of NC Systems
CNC machine tool systems can be classified in various ways such as : 1. Point-to-point or contouring : depending on whether the machine cuts metal while the workpiece moves relative to the tool 2. Incremental or absolute : depending on the type of coordinate system adopted to parameterise the motion commands 3. Open-loop or closed-loop : depending on the control system adopted for axis motion control

Point-to-point systems
Point-to-point (PTP) systems are the ones where, either the work piece or the cutting tool is moved with respect to the other as stationary until it arrives at the desired position and then the cutting tool performs the required task with the motion axes stationary. Such systems are used, typically, to perform hole operations such as drilling, boring, reaming, tapping and punching. In Version 2 EE IIT, Kharagpur 7

a PTP system, the path of the cutting tool and its feed rate while traveling from one point to the next are not significant, since, the tool is not cutting while there is motion. Therefore, such systems require only control of only the final position of the tool. The path from the starting point to the final position need not be controlled.

Contouring systems
In contouring systems, the tool is cutting while the axes of motion are moving, such as in a milling machine. All axes of motion might move simultaneously, each at a different velocity. When a nonlinear path is required, the axial velocity changes, even within the segment. For example, cutting a circular contour requires sinusoidal rates of change in both axes. The motion controller is therefore required to synchronize the axes of motion to generate a predetermined path, generally a line or a circular arc. A contouring system needs capability of controlling its drive motors independently at various speeds as the tool moves towards the specified position. This involves simultaneous motion control of two or more axes, which requires separate position and velocity loops. It also requires an interpolator program that generates the position and velocity setpoints for the two drive axes, continuously along the contour. In modern machines there is capability for programming machine axes, either as point-to-point or as continuous (that is contouring) Before the next type of classification is introduced, it is necessary to present the basic coordinate system conventions in a machine tool.

Point to Ponder: 2
A. Comment on the sensing requirements for PTP and Contouring axes B. Do you think the overall cutting time can be optimized for PTP and Contouring systems? Are there any constraints to that?

Coordinate Systems
The coordinate system is defined by the definition of the translational and rotational motion coordinates. Each translational axis of motion defines a direction in which the cutting tool moves relative to the work piece. The main three axes of motion are referred to as the X, Y. and Z axes. The Z axis is perpendicular to both X and Y in order to create a right-hand coordinate system, such as shown in Fig. 23.5. A positive motion in the Z direction moves the cutting tool away from the workpiece. The location of the origin is generally adjustible. Figure 23.4 shows the coordinate system for turning as in a lathe while Fig. 23.5 shows the system for drilling and milling. For a lathe, the infeed/radial axis is the x-axis, the carriage/length axis is the z-axis. There is no need for a y-axis because the tool moves in a plane through the rotational center of the work. Coordinates on the work piece shown below are relative to the work.

Version 2 EE IIT, Kharagpur 8

Head z

Tail Stock

x

y

Fig. 23.4 Co-ordinate system for turning In drilling and milling machines the X and Y axes are horizontal. For example, a positive motion command in the drill moves the X axis from left to right, the Y axis from front to back, and the Z axis toward the top. In the lathe only two axes are required to command the motions of the tool. Since the spindle is horizontal, the Z axis is horizontal as well. The cross axis is denoted by X. A positive position command moves the Z axis from left to right and the X axis from back to front in order to create the right-hand coordinate system.

z

y x

Fig. 23.5 Co-ordinate system for drilling and milling For a tool with a horizontal spindle the x-axis is across the table, the y-axis is down, and the zaxis is out. In addition to the translational motion, rotary motions around the axes parallel to X, Y, and Z can also be defined. Similarly, in addition to the primary motion coordinates, secondary coordinates can also exist.

Incremental Systems
In an incremental system the movements in each Part program block are expressed as the displacements along each coordinate axes with reference to the final position achieved at the end of executing the previous program block.

Version 2 EE IIT, Kharagpur 9

y F B A C E D

x 30 50 70 100 130

Fig. 23.6 A trajectory for drilling Consider, for example, the trajectory of rectilinear motions shown in Fig. 23.6 for a PTP system. In an incremental system, the motion parameters, along the X-axis, for the segments, A-B, B-C, C-D, D-E, E-F and F-A, would be given as, 50, 20, 60, -30, -70 and –30, respectively.

Absolute System
An absolute NC system is one in which all position coordinates are referred to one fixed origin called the zero point. The zero point may be defined at any suitable point within the limits of the machine tool table and can be redefined from time to time. Any particular definition of the zero point remains valid till another definition is made. In the Fig. 23.6, considering the X-coordinate for point A as zero, the X-coordinate for points B and C would be 50 and 70, respectively, in an absolute coordinate system. Most modem CNC systems permit application of both incremental and absolute programming methods. Even within a specific part program the method can be changed These CNC systems provide the user with the combined advantages of both methods

Unit of Displacement
Displacements are expressed in part programs by integers. Each unit corresponds to the position resolution of the axes of motion and will be referred to as the basic length-unit (BLU). The BLU is the smallest length of motion that can be repeatably sensed in the machine. It therefore determines the accuracy of machining possible with a given machine. For example, if a shaft angle encoder having a sensitivity of 500 pulses per revolution is mounted on a lead screw having a pitch of 5 mm, the BLU is 0.01 mm. In modern CNC machines the programmer can use floating-point dimensional data, which would be converted in to BLU’s by the interpreter.

Version 2 EE IIT, Kharagpur 10

Point to Ponder: 3
A. Can you think of one advantage and one disadvantage of the incremental coordinate system compared to the absolute one? B. Is there any connection between the choice of coordinate system and the position sensing machine used for the machine tool? C. How would you decide on the BLU for systems with position sensors such as LVDTs and resolvers? D. Is the BLU affected by the motor or the drive system also?

Part Programming
As mention earlier, a part program is a set of instructions often referred to as blocks, each of which refers to a segment of the machining operation performed by the machine tool. Each block may contain several code words in sequence. These provide: 1. Coordinate values (X, Y, Z, etc.) to specify the desired motion of a tool relative to a work piece. The coordinate values are specified within motion codeword and related interpolation parameters to indicate the type of motion required (e.g. point-to-point, or continuous straight or continuous circular) between the start and end coordinates. The CNC system computes the instantaneous motion command signals from these code words and applies them to drive units of the machine. 2. Machining parameters such as, feed rate, spindle speed, tool number, tool offset compensation parameters etc. 3. Codes for initiating machine tool functions like starting and stopping of the spindle, on/off control of coolant flow and optional stop. In addition to these coded functions, spindle speeds, feeds and the required tool numbers to perform machining in a desired sequence are also given. 4. Program execution control codes, such as block skip or end of block codes, block number etc. 5. Statements for configuring the subsystems on the machine tool such as programming the axes, configuring the data acquisition system etc. A typical block of a Part program is shown below in Fig. 23.7. Note that the block contains a variety of code words such G codes, M codes etc. Each of these code words configure a particular aspect of the machine, to be used during the machining of the particular segment that the block programmes.

Version 2 EE IIT, Kharagpur 11

N 1234 G-Address of block number Block number Preparatory function Transient information Spindle speed function Miscellaneous function Feed function Tool function Compensation function End of block

X-- Y--

S--

M-- F-- T-- D--

LF

Fig. 23.7 Structure of a block in a part program Appendix-1 provides some details of these codes. A typical sequence of operations in a part program would be, A. Introductory functions such as units, coordinate definitions, coordinate conventions, such as, absolute or relative etc. B. Feeds, speeds, etc. C. Coolants, doors, etc. D. Cutting tool movements and tool changes E. Shutdown

Point to Ponder: 4
A. Consider the part program segment given below.
N0010G90; N0011G01X1Y2; N0012G01X2Y2; N0013G91; N0014G01X1; N0015G92X2Y2; N0016G01X1Y1;

Identify the meaning of the codes B. Draw the trajectory of table motion that this program seeks to create. C. Consider the part program segment given below for cutting a circular arc. Version 2 EE IIT, Kharagpur 12

N10G01XY1; N11G03X2Y512J1;

Determine the parameters of the circle. D. Is there any other way of operating machine tools other than by Part programs? Before we discuss the next categorization of CNC systems, it is important to understand how the instructions in a part program are converted to the operations needed for machining the parts. The control system software, which controls the axis motion, is called the axis manager. The axis manager controls the movement of the axes on the machine tool. This control may be divided into two distinct activities, namely, Axes interpolation Axes servo control

These two activities are executed by two specific routines, namely the interpolation and servo control routines, which communicate by means of a buffer for the exchange of data. The axis manager is processed by one or more dedicated CPUs. In a multi-processor architecture, the interpolation and the servo control can be split between the various CPUs according to different combinations, such as, interpolation of all the axes on one CPU and servo control of all the axes on another CPU interpolation of all the axes on one CPU and servo control of part of the axes on the same CPU and servo control of the remaining axes on another CPU.

Interpolation
Interpolation consists in the calculation of the coordinated movement of several axes using the programmed parameters, in order to obtain a resulting trajectory, which can be of various types, such as: Straight line Circular Helicoidal

The interpolation module computes instant by instant position commands for the servo module, which in turn, drives the motors. There are two types of interpolators, namely: Process interpolator (for continuous axes) Point-to-point interpolator (for point-to-point axes)

Servo Control
Servo control consists of all the activities which allow several axes to effectively maintain the trajectory calculated by the interpolator. Continuous axes are continuously controlled by the system both for “speed” and for “position” so as to guarantee that the calculated trajectory is maintained. In contrast, for point-to-point axes there is no guarantee that the trajectory will be maintained. The only guarantee is that the final point will be reached.

Version 2 EE IIT, Kharagpur 13

Types of servo control for motion axes
The axes controlled by the axis manager may be divided into various types according to the specific function they perform on the machine tool. Some of these types are described below.

Coordinated Axis
This is a working axis, which may be interpolated along with other axes of the same type. This is necessary for generating specific 2D or 3D contours. The movement of one of the axes can be taken as the master and the other axes slaved to it. The mechanical and electrical features of the slave axis must be identical to those of the master. A coordinated axis can also be rotary and programmed in degrees. Note that for rotary axes, it may or may not be needed to map angular displacements to a (0-2π) interval.

Point-to-point Axis
This axis is not required to be interpolated with others, since it is used for only for positioning from one point to another. Such an axis may be viewed as an independent mechanical component fitted with a positioning transducer.

Spindle Axis
There are two types of spindle axes. For some, only the speed of the axis need to be controlled and not the position by the spindle servo control system. Such an axis essentially realizes a “motorized” tool. For the second type, the speed of this spindle axis, as well as its angular position can be controlled. This has application in controlling threading processes. It is also possible to drive the spindle in coordinated motion, interpolated with the other axes. This uses the spindle transducer value as the set point for the other axes. A typical example is the C axis in lathes. One can command a controlled acceleration ramp for the spindle rotation command. However, for improved angular positioning, this must be eliminated. It is also possible to have spindle drives without servo control, generally for spindles driven with ac motors. The only control needed in such a case is for reversal of spindle rotation. For control of tool and workpiece motion in the various ways described above, one of two kinds of control systems is employed.

Open Loop Systems
The term open-loop means that there is no feedback, and in open loop systems the motion controller produces outputs depending only on its set points, without feedback information about the effect that the output produces on the motion axes. We have already seen that the effects of controller outputs on the plant may not be the same always, since it depends on factors such as loads, parameter variations in the plant etc. In open loop systems, the set points are computed from the instructions in the Part program and fed to the controller, which may reside in a different microprocessor, through an interface. These motion commands may be in the form of electrical pulses (typical for step motor drives) or analog or digital signals, and converted to

Version 2 EE IIT, Kharagpur 14

speed or current set points by the controller. These setpoints, in turn, are sent to the power electronic drive system that applies the necessary voltage/current to the motors. The primary drawback of open-loop system is that there is no feedback system to check whether the commanded position and velocity has been achieved. If the system performance is affected by load, temperature or friction then the actual output could deviate from the desired output. For these reasons, the open-loop system is generally used in point-to-point systems where the accuracy requirements are not critical. Contouring systems do not use open-loop control.

Closed Loop systems
Closed-loop control, as described in the module on controllers, continuously senses the actual position and velocity of the axis, using digital sensors such as encoders or analog sensors such resolvers and tachogenerators and compares them with the setpoints. The difference between the actual value of the variable and its setpoint is the error. The control law takes the error as the input and drives the actuator, in this case the servo motor and its drive system, to achieve motion variables that are close to the set points. As we know, closed loop systems can achieve much closer tracking of set points even with disturbances and parameter variations in the system with, say, with temperature. Closed-looped systems, on the other hand, require more complex control as well as feedback devices and circuitry in order for them to implement both position and velocity control. Most modern closed-loop CNC systems are able to provide very close resolution of 0.0001 of an inch.

Point to Ponder: 5
A. Why is closed loop control required for continuous axes and not for PTP axes? B. What sort of motors, drives, controllers and sensors would you recommend for continuous axis control? C. Comment on the comparative requirements of computing speed and memory for PTP and continuous axis control

Lesson Summary
In this lesson, the following topics related to CNC machines have been discussed. A. Fundamentals of machining and its parameters B. Definition and advantages C. Introduction to coordinate systems D. Part programming E. Basic axis control strategies F. Classification

Version 2 EE IIT, Kharagpur 15

Appendix-1
In this appendix we provide a list of G and M-codes for the reader to have an idea of the kind of functionality that can be realized using these codes. These codes were originally designed to be read from paper tapes and are designed to direct tool motion with simple commands. A basic list of ‘G’ operation codes is given below. These direct motion of the tool. G00 - Rapid move (not cutting) G01 - Linear move G02 - Clockwise circular motion G03 - Counterclockwise circular motion G04 - Dwell G05 - Pause (for operator intervention) G08 - Acceleration G09 - Deceleration G17 - x-y plane for circular interpolation G18 - z-x plane for circular interpolation G19 - y-z plane for circular interpolation G20 - turning cycle or inch data specification G21 - thread cutting cycle or metric data specification G24 - face turning cycle G25 - wait for input to go low G26 - wait for input to go high G28 - return to reference point G29 - return from reference point G31 - Stop on input G33-35 - thread cutting functions G35 - wait for input to go low G36 - wait for input to go high G40 - cutter compensation cancel G41 - cutter compensation to the left G42 - cutter compensation to the right G43 - tool length compensation, positive G44 - tool length compensation, negative G50 - Preset position G70 - set inch based units or finishing cycle G71 - set metric units or stock removal G72 - indicate finishing cycle G72 - 3D circular interpolation clockwise G73 - turning cycle contour G73 - 3D circular interpolation counter clockwise G74 - facing cycle contour G74.1 - disable 360 deg arcs G75 - pattern repeating G75.1 - enable 360 degree arcs G76 - deep hole drilling, cut cycle in z-axis G77 - cut-in cycle in x-axis Version 2 EE IIT, Kharagpur 16

G78 - multiple threading cycle G80 - fixed cycle cancel G81-89 - fixed cycles specified by machine tool manufacturers G81 - drilling cycle G82 - straight drilling cycle with dwell G83 - drilling cycle G83 - peck drilling cycle G84 - taping cycle G85 - reaming cycle G85 - boring cycle G86 - boring with spindle off and dwell cycle G89 - boring cycle with dwell G90 - absolute dimension program G91 - incremental dimensions G92 - Spindle speed limit G93 - Coordinate system setting G94 - Feed rate in ipm G95 - Feed rate in ipr G96 - Surface cutting speed G97 - Rotational speed rpm G98 - withdraw the tool to the starting point or feed per minute G99 - withdraw the tool to a safe plane or feed per revolution G101 - Spline interpolation M-Codes control machine functions. M00 - program stop M01 - optional stop using stop button M02 - end of program M03 - spindle on CW M04 - spindle on CCW M05 - spindle off M06 - tool change M07 - flood with coolant M08 - mist with coolant M08 - turn on accessory (e.g. AC power outlet) M09 - coolant off M09 - turn off accessory M10 - turn on accessory M11 - turn off accessory or tool change M17 - subroutine end M20 - tailstock back M20 - Chain to next program M21 - tailstock forward M22 - Write current position to data file M25 - open chuck M25 - set output #1 off M26 - close chuck M26 - set output #1 on Version 2 EE IIT, Kharagpur 17

M30 - end of tape (rewind) M35 - set output #2 off M36 - set output #2 on M38 - put stepper motors on low power standby M47 - restart a program continuously, or a fixed number of times M71 - puff blowing on M72 - puff blowing off M96 - compensate for rounded external curves M97 - compensate for sharp external curves M98 - subprogram call M99 - return from subprogram, jump instruction M101 - move x-axis home M102 - move y-axis home M103 - move z-axis home Some other typical codes and keywords used in part programs are given below. Annn - an orientation, or second x-axis spline control point Bnnn - an orientation, or second y-axis spline control point Cnnn - an orientation, or second z-axis spline control point, or chamfer Fnnn - a feed value (in ipm or m/s, not ipr), or thread pitch Innn - x-axis center for circular interpolation, or first x-axis spline control point Jnnn - y-axis center for circular interpolation, or first y-axis spline control point Knnn - z-axis center for circular interpolation, or first z-axis spline control point Lnnn - arc angle, loop counter and program cycle counter Nnnn - a sequence/line number Onnn - subprogram block number Pnnn - subprogram reference number Rnnn - a clearance plane for tool movement, or arc radius, or taper value Qnnn - peck depth for pecking cycle Snnn - cutting speed (rpm), spindle speed Tnnn - a tool number Unnn - relative motion in x Vnnn - relative motion in y Wnnn - relative motion in z Xnnn - an x-axis value Ynnn - a y-axis value Znnn - a z-axis value ; - starts a comment , or end of block

Version 2 EE IIT, Kharagpur 18

Appendix-2 Typical Specifications of a CNC System
1. Number of controlled axes 2. Interpolation 3. Resolution 4. Feed rate 5. Rapid traverse rate 6. Operating modes 7. Type of feedback : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : Two/Four/Eight, etc. Linear/circular/parabolic or cubic/cylindrical Input resolution (feedback) Programming resolution Feed/min Feed/revolution Feed rate override Feed/min Manual/Automatic/MDI(editing)/Input/Output/ Machine data set-up/Incremental, etc. Digital (rotary encoders with train of pulses) Analog (transducers, etc.) Both Number of characters which can be stored Part program input devices Output devices Editing of part program Through MDI Graphic simulation Blue print programming Background editing Menu driven programming Conversational programming Backlash Lead screw pitch error Temperature Cutter radius compensation Tool length compensation Built-in (integrated)/External Type of communication with NC Number of inputs, outputs, timers, counters and flags User memory Program organization Programming Languages Types of threads that can be cut Analog/Digital control Spindle orientation Spindle speed overrides RPM/min; constant surface speed Inch/metric switchover Polar coordinate inputs Mirror imaging Scaling Version 2 EE IIT, Kharagpur 19

8. Part program handling

9. Part programming

10. Compensations

11. Programmable logic controller

12. Thread cutting/Tapping 13. Spindle control

14. Other features

: : : : : : : : : :

Coordinate rotation system Custom macros Built-in fixed cycles Background communication Safe zone programming Built-in diagnostics, safety function, etc. Number of universal interfaces Number of active serial interfaces Direct numerical control interface Network interface capability

Version 2 EE IIT, Kharagpur 20

Answers, Remarks and Hints to Points to Ponder Point to Ponder: 1
A. Consider the metal cutting configurations shown. Identify the motion directions corresponding to speed, feed and depth of cut Ans: The directions are shown below.
Cutting velocity

Depth of cut

Feed motion

B. “Metal cutting is essentially a shearing phenomenon”, justify the statement with respect to the entries in Table 23.1. Ans: All entries in the table would show that the motion of the tool and the motion of the work piece create a shearing.

Point to Ponder: 2
A. Comment on the sensing requirements for PTP and Contouring axes Answer: PTP systems require only feedback of position. Contouring axes require feedback of position and velocity both. In PTP systems velocity is varied in open loop to achieve rapid traversal of the table with preprogrammed acceleration, deceleration patterns. However, often the hardware capabilities of the axis conforms to contouring. B. Do you think the overall machining time can be optimized for PTP and Contouring systems? Are there any constraints to that? Ans: For PTP systems the rapid traverse feature saves overall machining time. This means that the table movements from point to point are carried out as fast as possible, with acceleration and deceleration features. Deceleration is needed, so that the final position is reached accurately and quickly. The extent to which time can be minimized depends on the motoring and braking torque levels of the table drive. For contouring systems the instantaneous ratios of velocities along the motion axes are important to maintain the contours. Therefore over all time has to be minimized maintaining them. Note that whenever the contours take sharp turns, the velocity ratios also do so. Therefore it is very difficult to change the ratios sharply, if the velocities themselves are high. Thus contour cutting around sharp edges have to be done at low speeds. This constrains the minimization of overall machining time to maintain dimensional accuracy. Version 2 EE IIT, Kharagpur 21

Point to Ponder: 3
A. Can you think of one advantage and one disadvantage of the incremental coordinate system compared to the absolute one? Ans: An advantage of the absolute system is that a change in any one a position coordinate in an instruction does not affect the rest of the part program. In the case of the incremental system, any such change would require corresponding changes to be made in all subsequent instructions in the part program. On the other hand, for incremental systems programming for parts with mirror image symmetry is easier, since it involves only changes in signs for the position commands with respect to the symmetrical points. Similarly, verification of program dimensions with drawing is easier with incremental system. B. Is there any connection between the choice of coordinate system and the position sensor used for the machine tool? Ans: Among the position sensors used, the most common ones are the linear optical scales or inductosyns which directly measure movement of the table slide with respect to the fixed parts of the machine. The other kinds are the rotary sensors such as the shaft angle encoders and resolvers. The translational motion of the slide is deduced from these using the ball screw pitch constant. Among these, the linear optical scale and the rotary encoder can either give absolute or incremental displacement, depending on the encoding on the grating. However, in fact incremental output is more common, since better resolution can be achieved. Thus, in an absolute coordinate system, this output must be integrated to generate absolute positions by electronic means. The resolver or the inductosyn, on the other hand naturally generate absolute positions with respect to a fixed origin. This means that these signals have to be differenced from the current origin of the incremental system, to generate incremental coordinate positions. C. How would you decide on the BLU for systems with continuous position sensors such as inductosyns and resolvers? Ans: D. Is the BLU affected by the motor or the drive system also?

Point to Ponder: 4
A. Explain the part program segment given below.
N0010G90; N0011G01X1Y2; N0012G01X2Y2; N0013G91; N0014G01X1; N0015G92X2Y2; N0016G01X1Y1;

Version 2 EE IIT, Kharagpur 22

Ans:
N0010G90; PUT IN ABSOLUTE MODE N0011G01X1Y2; MOVE TO (1, 2) N0012G01X2Y2; MOVE TO (2, 2) N0013G91; PUT IN INCREMENTAL MODE N0014G01X1; MOVE TO (3, 2) N0015G92X2Y2; SET NEW ORIGIN N0016G01X1Y1; MOVE TO (3, 3) ABSOLUTE N0017G92X0Y0Z0; RESET THE ZERO

B. Draw the trajectory of table motion that this program seeks to create. Ans: (3, 3) (2, 2) (3, 2) (0, 0) (1, 2)

Version 2 EE IIT, Kharagpur 23

Module 5
CNC Machines
Version 2 EE IIT, Kharagpur 1

Lesson 24
CNC Machines: Interpolation, Control and Drive
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
After learning the lesson students should be able to A. Define the major subsystems for motion control B. Describe the major features of an interpolator for a contouring CNC system C. Distinguish and compare open loop control and closed loop CNC D. Name desirable features of feed and spindle drives of CNC machines

Introduction
The most critical and specialised activity in a CNC is axis management, which involves interpolation, servo control and drive of the motion axes. Axis management tasks can be processed by one or more dedicated CPUs. Often, the interpolation and the servo control tasks for the several motion axes can be split between the various CPUs. Both point-to-point (PTP) interpolators and contouring interpolators are available on a machine. As we shall see below, interpolation for PTP axes is extremely simple, and involves only providing the final position coordinates to the control system. For contouring systems, however, the interpolator must cyclically compute motion set points. We describe the process of contour generation by interpolation first. Once the set points are generated, these are provided to the servo control loops that compute the control inputs, based on the set points and the motion feedbacks (for closed loop systems) and provide such control inputs to the motor drive system. Basic approaches for control are described next to the interpolation. Finally, the drive system is to receive these inputs, in analog or digital form and would compute quantities such as voltage and current references and apply to the motor. Such details of CNC drive systems are also presented

Contour Generation by Interpolation
In contouring systems the machining path is usually constructed from a combination of linear and circular segments. It is only necessary to specify the coordinates of the initial and final points of each segment, and the feed rate. The operation of producing the required shape based on this information is termed interpolation and the corresponding unit is the “interpolator”. The interpolator coordinates the motion along the machine axes, which are separately driven, by providing reference positions instant by instant for the position-and velocity-control loops, to generate the required machining path. Typical interpolators are capable of generating linear and circular paths. Basically there are two types of CNCs, the reference-pulse and the reference words systems. These two CNC types require distinct interpolation routines in the control program to generate their corresponding reference signals (pulses or binary words). In the reference-pulse system, the computer produces a sequence of reference pulses for each axis of motion, each pulse generating a motion of one BLU. The accumulated number of pulses represents position, and the pulse frequency is proportional to the axis velocity. These pulses can either actuate a stepping motor in an open loop system, or be fed as a reference to a closed-loop system. With the sampled-data technique the control loop of each axis is closed through the computer itself, which generates reference binary words.

Version 2 EE IIT, Kharagpur 3

Reference-pulse interpolators are simpler to program, but there is a restriction on the maximum axis velocity imposed by the interpolation execution time. Reference-pulse interpolators execute cyclically with a clock. The maximum axis velocity is proportional to the maximum attainable clock frequency, which, in turn, depends on the execution time of the interpolator algorithm. Consequently, reference-pulse interpolators are often written in assembly language to improve the computing speed so as to attain higher maximum attainable axis velocity.

Point to Ponder: 1
A. Is it possible to have a reference pulse interpolator for a CNC machine with dc drive? B. What limits the speed of operation in reference word interpolator based systems? Below we present a linear reference pulse interpolation technique. At the heart of the interpolator is a Digital Differential Analyser (DDA) algorithm.

DDA Algorithm
DDA is essentially an algorithm for digital integration and generates a pulse train varying in frequency. Digital integration is performed by successive additions using an Euler approximation method shown in Fig. 24.1. p p = f(t) Δp

Δt

t

Fig. 24.1 Backward Euler Integration From the above, let,
z ( t ) = ∫ p dt
t 0

The value of z at t = k∆t is denoted by zk, which may be written as: zk = zk −1 + Δzk where Δzk = pk Δt The value of pk can in turn be modified by incrementing or decrementing it by ∆p, which is either 1 or 0. The DDA integrator operates cyclically at a frequency f provided by an external clock. At each iteration the variable p is added to the register q so that, qk = qk −1 + pk At intervals this addition would generate an overflow bit, which is fed as the output reference pulse. Obviously, the higher the value of p the higher would be the frequency of generation of an Version 2 EE IIT, Kharagpur 4

overflow and a reference pulse. Thus the rate of generartion of the reference pulse would be proportional to the value of p. A schematic diagram of a DDA integrator is shown in Fig. 24.2. It consists of two n-bit registers, p and q, and one adder. Clock q(i +1) = q(i) + p q register q>L Adder Overflow bit up ∆p down + up down - counter (p) Yes q(i + 1) = q(i + 1) - L Generate Motion Pulse No

Fig. 24.2 DDA Integrator Schematics and Flowchart A symbolic representation of a DDA integrator is showing in Fig. 24.3. Δp Clock Fig. 24.3 A symbolic representation of a DDA integrator With the advent of fast microprocessors, the need for hardwired DDA’s have reduced. Instead, the DDA algorithm is implemented using the registers of the microprocessor, in software. + -

f0

Linear Reference Pulse Interpolation
The ability to control the movement along a straight line between given initial and final coordinates is termed linear interpolation. In this lesson only 2-D linear interpolators are discussed. A 2-D linear interpolator supplies velocity commands, in pulses per second, simultaneously to two machine axes, and adjusts the ratio between the pulse frequencies depending on the slope of the trajectory. For example, consider the case in Fig. 24.4, where a straight path has to be cut between points A and B. Note that movement along each axis takes place by 1 BLU for every reference pulse along the axis. The interpolator therefore has to provide pulses to each axis at definite rates (say, from Figure 24.4, a and b pulses per second, along X and y axes respectively) with respect to time.

Version 2 EE IIT, Kharagpur 5

Y

B

Desired path b Interpolated path a Fig. 24.4 2D linear interpolation A 2-D linear interpolator based on DDA integration is shown in Fig. 24.5. + 1 - p=a Δz X

A

Δt

X axis

Clock + 2 - p=b Δz

Δt

Y axis

Fig. 24.5 DDA-based Linear Reference Pulse Interpolator. A common clock controls both the integrators. The output pulses actuate stepping motors in open-loop systems, where each pulse causes a single step motion, or can be fed as reference to closed-loop systems. In addition to maintaining the proper velocity ratio between the two axes, a desired velocity along the path need also be maintained. This is achieved by controlling the clock frequency of the first two DDAs.

Reference-word Circular Interpolators
In reference pulse systems a pulse train of varying frequency is output to the servo control module. The servo system for an axis causes an incremental displacement along the axis, for each pulse. As mentioned before, this can cause a speed limitation for the CNC, depending on the execution speed of the interpolation loop. In contrast, in reference word interpolation systems the maximum velocity is not limited by the execution speed of the processor. . The interpolation subroutines continuously provide velocity set points to the servo system, which realizes it through the drive. In this lesson we discuss a circular interpolation using the reference word method. This require the use of a “controlled speed drive” rather then a “position servo”. In circular interpolation, at a constant tangential velocity, V and radius R, the axial velocities satisfy the following equations: Version 2 EE IIT, Kharagpur 6

Vx ( t ) = V sin θ ( t ) Vy ( t ) = V cos θ ( t ) The velocity components Vx and Vy are computed by the circular interpolator and are supplied as reference inputs to the computer closed loops. Actually what is generated is a polygon inscribed on a circle. At the beginning of each side the interpolator provides new velocity references to the axes. The more the number of sides of the polygon, the better is the accuracy of the generated circle. The optimal number of sides is the smallest one for which the path error is within one BLU.
y

(Xi + 1, Yi + 1) ∆Yi θi + 1 α θi x

∆Xi

(Xi, Yi)

Fig. 24.6 Position and velocities at two successive points on a circle From Fig. 24.6 one can derive the following recursive update eqns. for the two coordinate axes. X ( i + 1) = AX ( i ) − BY ( i )

Y ( i + 1) = AY ( i ) + BX ( i )
where, A = cos α and B = sin α . The velocity set points for the axis drives are computed as follows. ΔX ( i ) = X ( i + 1) − X ( i ) = ( A − 1) X ( i ) − BY ( i )

ΔY ( i ) = Y ( i + 1) − Y ( i ) = ( A − 1) Y ( i ) + BX ( i ) Vx ( i ) = K ΔX ( i ) Vy ( i ) = K ΔY ( i )
where K = V/Rα. These velocity set points are provided to the servo control systems which are described below.

Point to Ponder: 2
A. Devise a scheme for changing the feedrate in an interpolator. B. How does one choose the value of α for circular reference word interpolation?

Version 2 EE IIT, Kharagpur 7

Servo Control
Servo control consists of all the activities, which allow several axes to effectively maintain the trajectory calculated by the interpolator. In CNC systems the position and velocity of the machine tool axes must be controlled closely and in a coordinated manner. Each axis is separately driven and follows the command signal produced by the interpolator. The control system can be either open-loop (as in PTP systems) or closed-loop (as in contouring systems).

Control of PTP Systems
In PTP systems only the final position of the table is controlled and the trajectory of motion in between the final and initial points are of no concern, since the tool is not cutting metal during motion. Open-loop controls using stepping motors as the drive devices of the machine table can be utilized in on small-sized point-to-point systems in which the load torque is small and constant. To save machine time, the table travels at high velocities. However, in open loop control there is no feedback of the actual position of the table. Therefore, the velocity must be gradually reduced towards the end, to avoid overshooting the final position, due to the limited braking torque compared to the momentum of the drive system and the table at high speed. As already explained in Lesson 23, PTP systems can use incremental or absolute programming. In incremental point-to-point systems a counter is loaded with the incremental coordinate of the destination by the interpolator. In closed loop systems it is decremented by pulses from the encoder, which indicate actual axis motion. In open loop control it is decremented at a suitable rate, by a pulse generator, as the step motor has turns by one step angle for each pulse. The motor is decelerated based on the content of the counter, which represents the distance to the destination point. A block diagram of an open-loop point-to-point control system for incremental programming for a single axis is shown in Fig. 24.7. When the motor axis reaches the destination point, counter content is zero.
Desired position from interpolator

Deceleration Unit

Down counter

End of count Generator

Pulse generator

AND

Desired direction of rotation from interpolator

Drive unit

Step motor

Table

Fig. 24.7 Incremental open-loop control for PTP systems

Version 2 EE IIT, Kharagpur 8

In open loop control, it is implicitly assumed that the shaft rotates by 1 BLU for every command pulse applied to the drive. Thus the pulse generator frequency cannot be increased beyond a certain level, which depends on the load, since then the stepper motor would not be able to turn under the load with each pulse. To obviate this difficulty, feedback is used. The incremental change in the shaft position is taken from an incremental position sensor, such as an encoder, on the leadscrew, as shown in Fig. 24.8. The encoder
Desired position from interpolator

Down counter

End of count generator

Decelerations Unit

Drive Unit

Step motor Desired direction from interpolator

Table

Incremental Position Sensor

Fig. 24.8 Block diagram of closed-loop incremental PTP system pulses, which represent the actual motion, feed the down counter rather than the command pulses produced by the pulse generator in the open-loop control. The decelerator circuit slows down the motor before the target point in order to avoid overshoot. Note that, even in a closed loop system, to avoid errors due to backlash in gears, an overshoot is to be avoided. When the table is at a close distance of the target point, the table “creeps” toward the final point at very low velocity, before it stops. In absolute positioning systems utilizing an incremental feedback device, two alternative sequences of pulses from the incremental encoder, one for each direction of motion, feed the up and down inputs of a position counter. Thus, its contents are incremented for a rightward movement of the corresponding axis and are decremented for a leftward motion. The position counter value, therefore, indicate the actual absolute position of the axis. A command register is loaded with the required absolute destination position of the axis, by the interpolator. The subtractor unit indicates the instantaneous actual difference between the required and actual position, which is the distance to the target point. The subtractor output is the position error of the loop. Till the subtractor output is zero, pulses are fed through a deceleration circuit to the motor.

Control of Contouring Systems
In contouring systems the tool is cutting while the machine axes are moving. The contour of the part is determined by the ratio between the velocities, along the two axes. The control in contouring systems operates in closed loop. Therefore, a contouring system uses a cascade control structure involving an inner velocity loop and an outer position loop for each feed axis improved dynamic response. In such systems the interpolator generate reference signals (in form of a sequence pulses or position words) for each axis of motion, in a coordinated manner so that a desired contour is generated. Version 2 EE IIT, Kharagpur 9

Typical cascade control structure of contouring systems is shown in Fig. 24.9. It uses an inner velocity feedback loops incorporating a tachometer usually mounted directly on the motor shaft and an outer position feedback loop which is capable of measuring incremental (such as from an incremental encoder) or absolute angular position of the leadscrew shaft (such as a resolver or inductosyn).
Position Reference input + Speed error Speed Controller and drive

Position Controller

+ -

Motor

Table

Position Sensor

Speed sensor

Position Feedback

Fig. 24.9 Control loop of a contouring system In encoder-based systems each pulse indicates a motion of 1 BLU of axis travel. Therefore, the number of pulses over a period represents incremental change in position over the period and the encoder pulse frequency is proportional to the axis velocity. In such a system, fed from a reference pulse interpolator the comparison is done by an up-down counter which is fed by two sequences of pulses: reference pulses from the interpolator and feedback pulses generated by the encoder. The counter produces a number representing the instantaneous position error in pulse units. This number can be converted by the DAC and fed to an analog position control system. A typical electronic PLC function module board for CNC drives is described below.

Point to Ponder: 3
A. Do you think the closed loop PTP system control loop would work for arbitry axis velocities? Justify your answer. B. What type controllers would you prefer for a cascade position controller for a contouring system?

A Typical PLC-based Motion Control Board for CNC Drive
A PLC may be enhanced by a motion control or a position control module to control a CNC machine. A position control module is suited to the positioning operation requiring a high degree of accuracy with fast closed-loop position control of two axes. The module measures and processed the digital impulses of the position measuring system (actual values) and provides the drives with their respective rotational speed set point values for position to be reached. The functional parameters of the position control module can be set and stored in the on-board memory. The central controller provides the set point values, parameters and control commands and reads the actual values.

Version 2 EE IIT, Kharagpur 10

Position Control Module • Shaft encoder input • Onboard dedicated high speed CPU • Motor drive set point output • Digital IO • Programmer port • Set point from data bus

Fig. 24.10 A typical PLC Function module for Axis Control in a CNC The module contains an on-board micro-processor which controls the measuring and processing of the actual values and emits, according to the design function, two rotational speed set point values. On the module there are plug-in connectors for position encoders. Either an incremental or absolute coded position encoder can be connected. Provision may also exist to accommodate plug-in connections for the analog speed set point values, the digital inputs and outputs (possibly for limit switches), as well as programmer interfaces. The functions of the position control module are controlled through the microprocessor software on the board EPROM. The data for the individual operating mode (machine date, operation data, process data) are provided via the programmer interface and are store on the on-board RAM. If incremental encoders are used for the measuring of actual values, their pulses are counted up or down by a 16-bit counter, depending on direction of rotation of the motor and can be doubled or quadrupled. When an absolute position encoder is connected, a GRAY-code with a maximum of 20 bit is processed by the operation mode. The set point output can take place via analog voltage outputs with a typical resolution of ± 11 bit. To improve the resolution in the lower speed range, a reference voltage for the DAC, individually switched for each section, with output voltage ranges between 0 and ± 1.25 V, ± 10V etc., may be provided by the operating mode. Digital inputs (end limit switches, reference point, external stop) and digital outputs (loop controller release) are generally available for each axis. The instantaneous position of the two axes is determined from pulse generated by the position encoders. A position controller (P controller) calculates the current rotational speed set point from the difference between the actual position and its set point. This is output as an analog signal (+ 10 V) and is available for the speed controller of the static converter of the particular axis. Operator input is possible either from the programmable controller or from programmer via two interfaces. Position set points are calculated from user commands and the interpretation of traversing programs. Incremental position encoders can be connected to the module. Both axes can be controlled from a programmer unit. Set points, actual values of process parameters such as speed or position, error and other information is displayed on the screen. Appropriate communication software is required for using the programmer unit (and for programming) and downloading the program onto the on-board memory. The positioning module can then be operated and tested independently of the programmable controller.

Version 2 EE IIT, Kharagpur 11

Before the positioning module and the processor module of a PLC system are able to communicate, the appropriate standard function blocks and the standard function blocks for communications processors must be loaded into the processor.

Axis and Spindle Drives
The primary function of the drive is to cause motion of the controlled machine tool member (spindle, slide, etc.) to conform as closely as possible to the motion commands issued by the CNC system. In order to maintain a constant material removal rate, the spindle and the tool movements have to be coordinated such that the spindle has a constant power and the slide has a constant torque over varying speed. In order to ensure a high degree of consistency in production, variable speed drives are necessary. With the developments in power electronics and microprocessor systems, variable drive systems have been developed. These are smaller in size, very efficient, highly reliable and meet all the stringent demands of the modern automatic machine tools. A discussion of variable speed drives for AC and DC motors can be found in Lessons 33-35. For CNC, typically AC/DC Servo and Stepper motor drives are used. For spindle drive adjustable speed DC or Induction motor drives are used.

Spindle Drives
The requirements of a spindle drive are mainly to control the set speed accurately within a wide constant power band, in the face of torque disturbances occurring with variations in material hardness. Large speed ranges up to 10-20,000 rpm and 1:1000 rangeability is often needed. The dc spindle drives are commonly used in machine tools. However, with the advent of microprocessor-based ac frequency inverter, of late, the ac drives are being preferred to dc drives as they offer many advantages. One of the main advantages with the microprocessor-based frequency inverter is the possibility of using the spindle motor for C-axis applications for speed control in the range of 1:10,00,000 with positioning. High overload capacity is also needed for unintended overloads on the spindle, say, due to an inappropriate feed.

Feed Drives
A feed drive consists of a feed servomotor and an electronic controller. Unlike a spindle motor, the feed motor needs to operate with constant torque characteristics. The drive speed should be extremely variable with a speed range of at least 1: 20,000, which means that both at a maximum speed, say of 2-3000 rpm, and at a minimum speed of 0.1 rpm, the feed motor must run smoothly. Positioning resolution corresponding to angular rotations of angular minutes is needed with quick response four-quadrant operation that needs a high torque-inertia ratio. Also, in contouring operations, where a prescribed path has to be followed continuously, several feed drives have to work simultaneously. This requires a fast response and matched dynamic characteristics high bandwidth for different axes. Variable speed dc feed drives are very common in machine tools because of their simple control techniques. However, with the advent of the latest power electronic devices and control techniques ac feed drives are becoming popular due to certain advantages.

Version 2 EE IIT, Kharagpur 12

Point to Ponder: 4
A. For the motion control board: a. Where is the interpolator situated? b. Where are the position and velocity loops situated? B. Why should a feed drive operate in constant torque mode, while the spindle drive should operate in a constant power mode?

Lesson Summary
In this lesson, the following topics related to CNC machines have been discussed. A. Reference pulse and reference word interpolators B. Linear and Circular Interpolation C. Digital Integration with a DDA D. Open loop and closed loop control E. Control of PTP and Contouring Systems F. Characteristics of Feed and Spindle Drives

Version 2 EE IIT, Kharagpur 13

Answers, Remarks and Hints to Points to Ponder Point to Ponder: 1
A. Is it possible to have a reference pulse interpolator for a CNC machine with dc drive? Ans: Yes, it is. Consider a closed loop system with an incremental encoder feedback. Implement an up-down counter with the pulse train from reference pulse interpolator driving the count-up input and the pulse train from the encoder driving the count-down input. The counter value indicates the instantaneous position error which can be used to drive say an analog controller through a DAC. B. What limits the speed of operation in reference word interpolator based systems? Ans: The servo system dynamics limit it, rather than the interpolator loop execution time. This is especially true when one is cutting angles and corners.

Point to Ponder: 2
A. Devise a scheme for changing the feedrate in an interpolator. Ans: This can be done by generating the interpolator clock input using an DDA integrator as shown below. Note that by changing the constant F, the interpolator clock frequency can be changed. System Clock Wait to Interrupt q(i +1) = q(i) + F No

Interpolator Clock

q > qM Yes q(i +1) = q(i+ 1) – qM

B. How does one choose the value of α for circular reference word interpolation? Ans: The upper bound is decided by the maximum allowable deviation from a perfect circle. The lower bound is decided by loop execution speed of the interpolator.

Version 2 EE IIT, Kharagpur 14

Point to Ponder: 3
A. Do you think the closed loop PTP system control loop would work for arbitrary axis velocities? Justify your answer. Ans: May not be. One of the reasons being that with high axial velocities there would be position overshoots. However, unless the counter can represent negative position errors correctly correct positions would not be reached. B. What type controllers would you prefer for a cascade position controller for a contouring system? Ans: Typically P or PD controllers are used in the position loop, since there is already one speed to position integration built into the open loop dynamics. Inner velocity loop controllers are generally proportional (usually a servo amplifier).

Point to Ponder: 4
C. For the motion control board: a. Where is the interpolator situated? Ans: The interpolator must be situated external to the board, such as, in a PLC processor module. b. Where are the position and velocity loops situated? Ans: The position loop is implemented in the on-board software. Since the board provides velocity set points, it is assumed that the speed loop exists within the drive system. D. Why should a feed drive operate in constant torque mode, while the spindle drive should operate in a constant power mode? Ans: Because both torque on the feed drive motor and the power of the spindle motor can be shown to be roughly proportional to the material removal rate in machining. For a required degree of finish, a certain maximum material removal rate is possible. This is set through speed, feed and depth of cut settings. For all possible settings, to be able to obtain the best possible material removal rates, the feed and spindle motors should operate in their constant torque and constant power regions, respectively.

Version 2 EE IIT, Kharagpur 15

Module 6
Actuators
Version 2 EE IIT, Kharagpur 1

Lesson 25
Control Valves
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
At the end of this lesson, the student should be able to: • • • • • • • • Explain the basic principle of operation of a pneumatically actuated control valve Distinguish between air-to-open and air-to-close valves. Explain the constructions and relative advantages and disadvantages of single- seated and double-seated valves. Name three types of control valves and sketch their ideal flow characteristics. Sketch the shapes of the plugs for three different types of control valves. Define the term rangeability. Explain the different between ideal and effective characteristics. Explain the advantage of using equal percentage valve over using linear control valve.

Introduction
The control action in any control loop system, is executed by the final control element. The most common type of final control element used in chemical and other process control is the control valve. A control valve is normally driven by a diaphragm type pneumatic actuator that throttles the flow of the manipulating variable for obtaining the desired control action. A control valve essentially consists of a plug and a stem. The stem can be raised or lowered by air pressure and the plug changes the effective area of an orifice in the flow path. A typical control valve action can be explained using Fig. 1. When the air pressure increases, the downward force of the diaphragm moves the stem downward against the spring.

Classifications
Control valves are available in different types and shapes. They can be classified in different ways; based on: (a) action, (b) number of plugs, and (c) flow characteristics.

Version 2 EE IIT, Kharagpur 3

Air

Diaphragm Spring

Stem

Flow

Plug Fig. 1 Control valve (a) Action: Control valves operated through pneumatic actuators can be either (i) air to open, or (ii) air to close. They are designed such that if the air supply fails, the control valve will be either fully open, or fully closed, depending upon the safety requirement of the process. For example, if the valve is used to control steam or fuel flow, the valve should be shut off completely in case of air failure. On the other hand, if the valve is handling cooling water to a reactor, the flow should be maximum in case of emergency. The schematic arrangements of these two actions are shown in Fig. 2. Valve A are air to close type, indicating, if the air fails, the valve will be fully open. Opposite is the case for valve B.

A • •

B

Fail open or Air to close : A Fail closed or Air to open : B

Fig. 2 Air to open and Air to close valves (b) Number of plugs: Control valves can also be characterized in terms of the number of plugs present, as single-seated valve and double-seated valve. The difference in construction between a single seated and double-seated valve are illustrated in Fig. 3. Version 2 EE IIT, Kharagpur 4

Referring Fig.1 (and also Fig. 3(a)), only one plug is present in the control valve, so it is single seated valve. The advantage of this type of valve is that, it can be fully closed and flow variation from 0 to 100% can be achieved. But looking at its construction, due to the pressure drop across the orifice a large upward force is present in the orifice area, and as a result, the force required to move the valve against this upward thrust is also large. Thus this type of valves is more suitable for small flow rates. On the other hand, there are two plugs in a double-seated valve; flow moves upward in one orifice area, and downward in the other orifice. The resultant upward or downward thrust is almost zero. As a result, the force required to move a double-seated valve is comparatively much less. But the double-seated valve suffers from one disadvantage. The flow cannot be shut off completely, because of the differential temperature expansion of the stem and the valve seat. If one plug is tightly closed, there is usually a small gap between the other plug and its seat. Thus, single-seated valves are recommended for when the valves are required to be shut off completely. But there are many processes, where the valve used is not expected to operate near shut off position. For this condition, double-seated valves are recommended. Stem Packing gland Stem Packing gland

Flow Plug Flow Plug

Single-seated control valve

Double-seated control valve

(a)
Fig. 3 Single-seated and double-seated valves

(b)

(c) Flow Characteristics: It describes how the flow rate changes with the movement or lift of the stem. The shape of the plug primarily decides the flow characteristics. However, the design of the shape of a control valve and its shape requires further discussions. The flow characteristic of a valve is normally defined in terms of (a) inherent characteristics and (b) effective characteristics. An inherent characteristic is the ideal flow characteristics of a control valve and is decided by the shape and size of the plug. On the other hand, when the valve is connected to a pipeline, its overall performance is decided by its effective characteristic.

Version 2 EE IIT, Kharagpur 5

Ideal Characteristics
The control valve acts like an orifice and the position of the plug decides the area of opening of the orifice. Recall that the flow rate through an orifice can be expressed in terms of the upstream and downstream static pressure heads as:

q = K1a 2 g (h1 − h2 ) (1) 3 where q = flow rate in m /sec. K1 = flow coefficient a = area of the control valve opening in m2 h1 = upstream static head of the fluid in m h2 = downstream static head of the fluid in m g = acceleration due to gravity in m/sec2. Now the area of the control valve opening (a) is again dependent on the stem position, or the lift. So if the upstream and downstream static pressure heads are somehow maintained constant, then the flow rate is a function of the lift (z), i.e. (2) q = f ( z) The shape of the plug decides, how the flow rate changes with the stem movement, or lift; and the characteristics of q vs. z is known as the inherent characteristics of the valve.
Let us define

qmax zmax where, qmax is the maximum flow rate, when the valve is fully open and zmax is the corresponding maximum lift. So eqn. (2) can be rewritten in terms of m and x as: (3) m = f ( x) and the valve sensitivity is defined as dm dx , or the slope of the curve m vs. x. In this way, the control valves can be classified in terms of their m vs. x characteristics, and three types of control valves are normally in use. They are: (a) Quick opening (b) Linear (c) Equal Percentage. The characteristics of these control valves are shown in Fig. 4. It has to be kept in mind that all the characteristics are to be determined after maintaining constant pressure difference across the valve as shown in Fig.4.

m=

q

and

x=

z

Version 2 EE IIT, Kharagpur 6

% flow (m) 100 80 60 40 20 0 0 20 40 60 80 100 Quick Opening Linear

ΔP is Constant ΔP

Equal Percentage % lift (x)

Fig. 4 Flow characteristics of control valves Different flow characteristics can be obtained by properly shaping the plugs. Typical shapes of the three types of valves are shown in Fig. 5

Equal percentage

Linear

Quick opening

Fig. 5 Valve plug shapes for the three common flow characteristics. For a linear valve, dm dx = 1 , as evident from Fig.5 and the flow characteristics is linear throughout the operating range. On the other hand, for an equal percentage valve, the flow characteristics is mathematically expressed as: dm (4) =βm dx where β is a constant. The above expression indicates, that the slop of the flow characteristics is proportional to the present flow rate, justifying the term equal percentage. This flow characteristics is linear on a semilog graph paper. The minimum flow rate m0 (flow rate at x=0) is never zero for an equal percentage valve and m can be expressed as: m = m0 e β x (5) Rangeability of a control valve is defined as the ratio of the maximum controllable flow and the minimum controllable flow. Thus:
Rangeability = maximum controllable flow minimum controllable flow

Rangeability of a control valve is normally in between 20 and 70. Version 2 EE IIT, Kharagpur 7

Effective Characteristics
So far we have discussed about the ideal characteristics of a control valve. It is decided by the shape of the plug, and the pressure drop across the valve is assumed to be held constant. But in practice, the control valve is installed in conjunction with other equipment, such as heat exchanger, pipeline, orifice, pump etc. The elements will have their own flow vs. pressure characteristics and cause additional frictional loss in the system and the effective characteristics of the valve will be different from the ideal characteristics. In order to explain the deviation, let us consider a control valve connected with a pipeline of length L in between two tanks, as shown in Fig. 6. We consider the tanks are large enough so that the heads of the two tanks H0 and H2 can be assumed to be constant. We also assume that the ideal characteristic of the control valve is linear. From eqn. (1), we can write for a linear valve: K1a = Kz where K is a constant and z is the stem position or lift. Now the pipeline will experience some head loss that is again dependent on the velocity of the fluid. Ho H2

L

Flow Fig. 6 Effect of friction loss in pipeline for a control valve The head loss ΔhL will affect the overall flow rate q and eqn.(1) can be rewritten as: (6) q = ⎡ K 2 g ( H 0 − H 2 − ΔhL ⎤ z ⎣ ⎦ The head loss (in m) can be calculated from the relationship: L v2 ΔhL = F (7) D 2g

where F = Friction coefficient L = Length of the pipeline in m D = inside diameter of the pipeline in m v = velocity of the flow in m. Further, the velocity of the fluid can be related to the fluid flow q (in m3/sec) as: q (8) v= π 2 D 4 Combining (7) and (8), we can write:

Version 2 EE IIT, Kharagpur 8

8 FL 2 q π 2 gD5 Substituting (9) in (6) and further simplifying, one can obtain: ⎡ 2g (H0 − H 2 ) ⎤ q = ⎢K ⎥z 1+ α z2 ⎦ ⎣ ΔhL =
where α =

(9)

(10)

16FLK 2 π 2 D5 From (10), it can be concluded that q is no longer linearly proportional to stem lift z, though the ideal characteristics of the valve is linear. This nonlinearity of the characteristics is dependent on the diameter of the pipeline D; i.e. smaller the pipe diameter, larger is the value of α and more is the nonlinearity. The nonlinearity of the effective valve characteristics can be plotted as shown in Fig. 7.

qmax Flow (m3/sec)

ideal characteristics

decreasing pipe diameters O

zmax Lift z (m)

Fig. 7 Effect of pipeline diameter on the effective flow characteristics of the control valve

The nonlinearity introduced in the effective characteristics can be reduced by mainly (i) increasing the line diameter, thus reducing the head loss, (ii) increasing the pressure of the source H0, (iii) decreasing the pressure at the termination H2. The effective characteristics of the control valve shown in Fig.7 are in terms of absolute flow q ) in eqn. (3) rate. If we want to express the effective characteristics in terms of m (= qmax deviation from the ideal characteristics will also be observed. Linear valve characteristics will deviate upwards, as shown in Fig. 8. An equal percentage valve characteristic will also shift upward from its ideal characteristic; thus giving a better linear response in the actual case.

Version 2 EE IIT, Kharagpur 9

100 % Flow (m)

effective characteristics

50 ideal characteristics 0

0

50 % Lift (x)

100

Fig. 8 Comparison of ideal and effective characteristics for a linear valve

Thus linear valves are recommended when pressure drop across the control valve is expected to be fairly constant. On the other hand, equal percentage valves are recommended when the pressure drop across the control valve would not be constant due to the presence of series resistance in the line. As the line loss increases, the effective characteristics of the equal percentage valve will move closer to the linear relationship in m vs. x characteristics.

Conclusion
A control valve is the final control element in a process control. Thus the effectiveness of any control scheme depends heavily on the performance of the control valve. The proper design and fabrication of the valve is very important in order to achieve the desired performance level. Moreover control valves are of different size and shapes. Only few types of control valves have been discussed here, leaving a large varieties of valves, those are in use, to name a few, globe valves, butterfly valves, V-port valves etc. We have discussed here the pneumatically actuated control valves, though electrically and hydraulically actuated valves are also not uncommon. The shape of a valve plug is not the only deciding factor for determining its effective flow characteristics, but other equipment connected in the line along with the control valve, also affect its flow characteristics. Thus the effective flow characteristics of a linear valve may become nonlinear, as has been shown in this lesson. For this reason, equal percentage valves are preferred in many cases, since their effective characteristics tend to be linear, in presence of head loss in the pipeline. There are distinct guidelines for selecting the valve size and shape depending on load change, pipeline diameter etc. Bypass lines are sometimes used with a control valve in order to change the flow characteristics of the valve.

References
1. 2. 3. 4. 5. D.R. Coughanowr: Process systems analysis and control (2/e), McgrawHill, NY, 1991. D.P. Eckman: Automatic process Control, Wiley Eastern, New Delhi, 1958. B. Liptak: Process Control: Instrument Engineers Handbook W.L. Luyben and M.L. Luyben: Essentials of Process Control, McgrawHill, NY, 1997. P. Harriott: Process Control, Tata-McGrawHill, New Delhi, 1991. Version 2 EE IIT, Kharagpur 10

Review Questions
1. 2. 3. 4. 5. 6. 7. 8. 9. Sketch the construction of a pneumatically actuated diaphragm type single-seated control valve. Discuss the construction, advantages and disadvantages of a double-seated control valve. When would you recommend to use an air-to-close control valve? Give an example. Sketch and discuss the plug shapes and ideal flow characteristics of three different types of control valves. Discuss the ideal flow characteristics of an equal percentage valve. Define the term rangeability of a control valve. Why is the property important? How does the friction loss of a pipeline connecting the control valve affect the flow characteristics of the valve? Explain clearly. Distinguish between the terms- ideal characteristics and effective characteristics. What is the advantage of using a equal percentage valve over a linear valve?

Version 2 EE IIT, Kharagpur 11

Module 6
Actuators
Version 2 EE IIT, Kharagpur 1

Lesson 26
Hydraulic Actuation Systems - I: Principle and Components
Version 2 EE IIT, Kharagpur 2

Lesson Objectives
After learning the lesson students should be able to • • • • Describe the principles of operation of hydraulic systems and understand its advantages Be familiar with basic hydraulic components and their roles in the system Describe the constructional and functional aspects of hydraulic pumps and motors Draw the graphical symbols used to depict typical hydraulic system components

Introduction
Hydraulic Actuators, as used in industrial process control, employ hydraulic pressure to drive an output member. These are used where high speed and large forces are required. The fluid used in hydraulic actuator is highly incompressible so that pressure applied can be transmitted instantaneously to the member attached to it. It was not, however, until the 17th century that the branch of hydraulics with which we are to be concerned first came into use. Based upon a principle discovered by the French scientist Pascal, it relates to the use of confined fluids in transmitting power, multiplying force and modifying motions. Then, in the early stages of the industrial revolution, a British mechanic named Joseph Bramah utilized Pascal’s discovery in developing a hydraulic press. Bramah decided that, if a small force on a small area would create a proportionally larger force on a larger area, the only limit to the force a machine can exert is the area to which the pressure is applied.

Principle Used in Hydraulic Actuator System Pascal’s Law
Pressure applied to a confined fluid at any point is transmitted undiminished and equally throughout the fluid in all directions and acts upon every part of the confining vessel at right angles to its interior surfaces.

Amplification of Force
Since pressure P applied on an area A gives rise to a force F, given as, F = P×A Thus, if a force is applied over a small area to cause a pressure P in a confined fluid, the force generated on a larger area can be made many times larger than the applied force that crated the pressure. This principle is used in various hydraulic devices to such hydraulic press to generate very high forces.

Version 2 EE IIT, Kharagpur 3

Conservation of Energy
Since energy or power is always conserved, amplification in force must result in reduction of the fluid velocity. Indeed if the resultant force is applied over a larger area then a unit displacement of the area would cause a larger volumetric displacement than a unit displacement of the small area through which the generating force is applied. Thus, what is gained in force must be sacrificed in distance or speed and power would be conserved.

Pump PA Q L Travel/unit time F

Fig. 26.1 Major hydraulic and mechanical variables

Point to Ponder: 1
A. Can you give an analogy of the force amplification in hydraulic system from an electrical system? B. Can you imagine what would happen, if the cylinder piston in Fig. 26.1 is stopped forcefully?

Advantages of Hydraulic Actuation Systems
Hydraulics refers to the means and mechanisms of transmitting power through liquids. The original power source for the hydraulic system is a prime mover such as an electric motor or an engine which drives the pump. However, the mechanical equipment cannot be coupled directly to the prime mover because the required control over the motion, necessary for industrial operations cannot be achieved. In terms of these Hydraulic Actuation Systems offer unique advantages, as given below. Variable Speed and Direction: Most large electric motors run at adjustable, but constant speeds. It is also the case for engines. The actuator (linear or rotary) of a hydraulic system, however, can be driven at speeds that vary by large amounts and fast, by varying the pump delivery or using a flow control valve. In addition, a hydraulic actuator can be reversed instantly while in full motion without damage. This is not possible for most other prime movers. Version 2 EE IIT, Kharagpur 4

Power-to-weight ratio: Hydraulic components, because of their high speed and pressure capabilities, can provide high power output with vary small weight and size, say, in comparison to electric system components. Note that in electric components, the size of equipment is mostly limited by the magnetic saturation limit of the iron. It is one of the reasons that hydraulic equipment finds wide usage in aircrafts, where dead-weight must be reduced to a minimum. Stall Condition and Overload Protection: A hydraulic actuator can be stalled without damage when overloaded, and will start up immediately when the load is reduced. The pressure relief valve in a hydraulic system protects it from overload damage. During stall, or when the load pressure exceeds the valve setting, pump delivery is directed to tank with definite limits to torque or force output. The only loss encountered is in terms of pump energy. On the contrary, stalling an electric motor is likely to cause damage. Likewise, engines cannot be stalled without the necessity for restarting.

Point to Ponder: 2
A. Consider two types of variable speed drives. In the first one an electric motor with a power electronic servo drive is directly coupled to the load through a mechanism. In the second one an electric motor with a constant speed drive drives the pump in a hydraulic system which provides the variable speed drive to the load. Which one of these two is more energy efficient? B. Why is stalling an electric motor is likely to cause damage? What can be done to prevent it?

Components of Hydraulic Actuation Systems Hydraulic Fluid
Hydraulic fluid must be essentially non-compressible to be able to transmit power instantaneously from one part of the system to another. At the same time, it should lubricate the moving parts to reduce friction loss and cool the components so that the heat generated does not lead to fire hazards. It also helps in removing the contaminants to filter. The most common liquid used in hydraulic systems is petroleum oil because it is only very slightly compressible. The other desirable property of oil is its lubricating ability. Finally, often, the fluid also acts as a seal against leakage inside a hydraulic component. The degree of closeness of the mechanical fit and the oil viscosity determines leakage rate. Figure 26.2 below shows the role played by hydraulic fluid films in lubrication and sealing. Film of hydraulic fluid lubricates

Film of hydraulic fluid seals passage from adjacent Fig. 26.2 Lubrication and Sealing by Hydraulic Fluid Version 2 EE IIT, Kharagpur 5

The Fluid Delivery Subsystem
It consists of the components that hold and carry the fluid from the pump to the actuator. It is made up of the following components.

Reservoir
It holds the hydraulic fluid to be circulated and allows air entrapped in the fluid to escape. This is an important feature as the bulk modulus of the oil, which determines the stiffness of hydraulic system, deteriorates considerably in the presence of entrapped air bubbles. It also helps in dissipating heat. From hydraulic system To hydraulic system PUMP Releases Bubbles

Dissipates Heat Fig. 26.3 The functions of the reservoir

Reservoir Baffle

Filter
The hydraulic fluid is kept clean in the system with the help of filters and strainers. It removes minute particles from the fluid, which can cause blocking of the orifices of servo-valves or cause jamming of spools.

Point to Ponder: 3
A. What would happen if orifices of valves are blocked by, say, a metal chip in the hydraulic oil?

Line
Pipe, tubes and hoses, along with the fittings or connectors, constitute the conducting lines that carry hydraulic fluid between components. Lines are one of the disadvantages of hydraulic system that we need to pay in return of higher power to weight ratio. Lines convey the fluid and also dissipate heat. In contrast, for Pneumatic Systems, no return path for the fluid, which is air, is needed, since it can be directly released into the atmosphere. There are various kinds of lines in a hydraulic system. The working lines carry the fluid that delivers the main pump power to the load. The pilot lines carry fluid that transmit controlling pressures to various directional and relief valves for remote operation or safety. Lastly there are drain lines that carry the fluid that inevitably leaks out, to the tank.

Version 2 EE IIT, Kharagpur 6

Working line


Pilot line Drain line

Fig. 26.4 The various kinds of lines in a hydraulic system Fig 26.5 below shows a typical configuration of connecting the supply and the return lines as well as the filter to the reservoir. The graphical symbol for a Reservoir and Filters is shown in Fig. 26.6. Supply Line

Pump Return Line

Filter Reservoir Fig. 26.5 Connection Arrangement of Filter and Lines with a Reservoir

Fig. 26.6 The graphical symbol for Reservoirs and Filters

Version 2 EE IIT, Kharagpur 7

Fittings and Seals
Various additional components are needed to join pipe or tube sections, create bends and also to prevent internal and external leakage in hydraulic systems. Although some amount of internal leakage is built-in, to provide lubrication, excessive internal leakage causes loss of pump power since high pressure fluid returns to the tank, without doing useful work. External leakage, on the other hand, causes loss of fluid and can create fire hazards, as well as fluid contamination. Various kinds of sealing components are employed in hydraulic systems to prevent leakage. A typical such component, known as the O-ring is shown below in Fig. 26.7.

O-Ring

Fig. 26.7 Sealing by O-rings

Hydraulic Pumps
The pump converts the mechanical energy of its prime-mover to hydraulic energy by delivering a given quantity of hydraulic fluid at high pressure into the system. Generically, all pumps are divided into two categories, namely, hydrodynamic or non-positive displacement and hydrostatic or positive displacement. Hydraulic systems generally employ positive displacement pumps only. The symbol for a pump, is shown in Fig. 26.8 below.

Pump Reversible Fig. 26.8 The graphical symbol for Pumps

Hydrostatic or Positive Displacement Pumps
These pumps deliver a given amount of fluid for each cycle of motion, that is, stroke or revolution. Their output in terms of the volume flow rate is solely dependent on the speed of the prime-mover and is independent of outlet pressure notwithstanding leakage. These pumps are generally rated by their volume flow rate output at a given drive speed and by their maximum operating pressure capability which is specified based on factors of safety and operating life considerations. In theory, a pump delivers an amount of fluid equal to its displacement each cycle or revolution. In reality, the actual output is reduced because of internal leakage or slippage which increases with operating pressure. Moreover, note that the power requirement on the prime mover theoretically increases with the pump delivery at a constant fluid pressure. If this power exceeds the power that the prime mover can handle the pump speed and the delivery rate would fall automatically. There are various types of pumps used in hydraulic systems as described below. Version 2 EE IIT, Kharagpur 8

Gear Pumps
Outlet

Drive Gear

Free Gear

Inlet Fig. 26.9 The construction of a Gear Pump A gear pump develops flow by carrying fluid between the teeth of two meshed gears. One gear is driven by the drive shaft and turns the other, which is free. The pumping chambers formed between the gear teeth are enclosed by the pump housing and the side plates. A low pressure region is created at the inlet as the gear teeth separate. As a result, fluid flows in and is carried around by the gears. As the teeth mesh again at the outlet, high pressure is created and the fluid is forced out. Figure 26.9 shows the construction of a a typical internal gears pump; Most gear type pumps are fixed displacement. They range in output from very low to high volume. They usually operate at comparatively low pressure.

Point to Ponder: 4
A. Why do gear pumps usually operate at comparatively low pressure?

Vane Pumps
In a vane pump a rotor is coupled to the drive shaft and turns inside a cam ring. Vanes are fitted to the rotor slots and follow the inner surface of the ring as the rotor turns (see Fig. 26.10). Centrifugal force and pressure under the vanes keep them pressed against the ring. Pumping chambers are formed between the vanes and are enclosed by the rotor, ring and two side plates. At the pump inlet, a low pressure region is created as the space between the rotor and ring increases. Oil entering here is trapped in the pumping chambers and then is pushed into the outlet as the space decreases.

Version 2 EE IIT, Kharagpur 9

Vane

System Pressure

Drive Shaft

Rotor

Fig. 26.10 Principle of Operation of Vane Pumps Most fixed displacement vane pumps today utilize the balanced design shown in Fig. 26.11. In this design, the cam ring is elliptical rather than a circle and permits two sets of internal ports. The two outlet ports are 180 degrees apart so that pressure forces on the rotor are cancelled out preventing side loading of the drive shaft and bearings Outlet Sense of rotation

Vane

. Inlet

Fig. 26.11 Construction of Balanced Vane Pumps

Piston Pumps
In a piston pumps, a piston reciprocating in a bore draws in fluid as it is retracted and expels it on the forward stroke. Two basic types of piston pumps are radial and axial. A radial pump has the pistons arranged radially in a cylinder block (shown in Fig. 26.12) in an axial pump the pistons are parallel to the axis of the cylinder block (shown in Fig. 26.13). The latter may be further divided into in-line (swash plate or wobble plate) and bent axis types.

Version 2 EE IIT, Kharagpur 10

Radial Piston Pumps
In a radial pump the cylinder block rotates on a stationary pintle and inside a circular reaction ring or rotor. As the block rotates, due to centrifugal force, charging pressure or some form of mechanical action the pistons remain pressed against the inner surface of the ring which is offset from the centerline of the cylinder block. Due the ring being off-centre, as the pistons reciprocate in their bores, they take in fluid as they move outward and discharge it as they move in. Centerline Cylinder Block Centerline Outlet Case Pintle Pistons

Inlet Cylinder Block Reaction Ring

Fig. 26.12 Cross Sectional View of Radial Piston Pumps

Swash Plate Design Inline Piston Pumps
In axial piston pumps, the cylinder block and drive shaft are co-axial and the pistons move parallel to the drive shaft. The simplest type of axial piston pump is the swash plate inline design shown in Fig. 26.13 and 26.14. The cylinder block in this pump is turned by the prime mover connected to the drive shaft. Pistons fitted to bores in the cylinder are connected to an angled swash plate. As the block turns, the piston shoes follow the swash plate, causing the pistons to reciprocate, since the distance of point of connection changes cyclically as the swash plate rotates. The fluid ports are placed in the valve plate so that the pistons pass the inlet port as they are being pulled out, so that fluid enters the cylinder cavity, and pass the outlet as they are being forced back in, delivering fluid into the system.

Version 2 EE IIT, Kharagpur 11

Stroke Length

Determines Swash Plate Angle that (Maximum Displacement) Fig. 26.13 Cross Sectional View of an Axial Piston Pump Valve Plate Slot Piston Sub – Assembly

Outlet Port.

Inlet Port.

Drive Shaft Swash Plate Shoe Plate (Retractor Ring).

Cylinder Block Bore.

Fig. 26.14 Cut-out View of Axial Piston Pump

Motors
Motors work exactly on the reverse principle of pumps. In motors fluid is forced into the motor from pump outlets at high pressure. This fluid pressure creates the motion of the motor shaft and finally go out through the motor outlet port and return to tank. All three variants of motors, already described for pumps, namely Gear Motors, Vane Motors and Piston motors are in use.

Accumulators
Unlike gases the fluids used in hydraulic systems cannot be compressed and stored to cater to sudden demands of high flow rates that cannot be supplied by the pump. An accumulator in a Version 2 EE IIT, Kharagpur 12

hydraulic system provides a means of storing these incompressible fluids under pressure created either by a spring, compressed a gas. Any tendency for pressure to drop at the inlet causes the spring or the gas to force the fluid back out, supplying the demand for flow rate.

Spring-Loaded Accumulators
In a spring loaded accumulator (Fig. 26.15), pressure is applied to the fluid by a coil spring behind the accumulator piston. The pressure is equal to the instantaneous spring force divided by the piston area. The pressure therefore is not constant since the spring force increases as fluid enters the chamber and decreases as it is discharged. Spring loaded accumulators can be mounted in any position. The spring force, i.e., the pressure range is not easily adjusted, and where large quantities of fluid are spring size has to be very large. Spring

Piston

Port

Fig. 26.15 A spring-loaded accumulator

Gas Charged Accumulator
The most commonly used accumulator is one in which the chamber is pre-charged with an inert gas, such as dry nitrogen. A gas charged accumulator should be pre-charged while empty of hydraulic fluid. Accumulator pressure varies in proportion to the compression of the gas, increasing as pumped in and decreasing as it is expelled.

Version 2 EE IIT, Kharagpur 13

Gas under pressure

Hydraulic Fluid

Fig. 26.16 A gas-charged accumulator

Cylinders
Cylinders are linear actuators, that is, they produce straight-line motion and/or force. Cylinders are classified as single-or double-acting as illustrated in Figures 26.17 and 26.18 with the graphical symbol for each type. Single Acting Cylinder: It has only one fluid chamber and exerts force in only one direction. When mounted vertically, they often retract by the force of gravity on the load. Ram type cylinders are used in elevators, hydraulic jacks and hoists.

Load Symbol

Load

From Pump

To Tank

Extend

Retract

Fig. 26.17 Cross Sectional View of Single-acting Cylinder Double-Acting Cylinder: The double-acting cylinder is operated by hydraulic fluid in both directions and is capable of a power stroke either way. In single rod double-acting cylinder there are unequal areas exposed to pressure during the forward and return movements due to the crosssectional area of the rod. The extending stroke is slower, but capable of exerting a greater force than when the piston and rod are being retracted. Version 2 EE IIT, Kharagpur 14

Load

Load

Exhaust To Tank From Pump

From Pump Exhaust To Tank

Extend Cylinder

Retract Cylinder

Fig. 26.18 Cross Sectional View of Single-acting Cylinder Double-rod double-acting cylinders are used where it is advantageous to couple a load to each end, or where equal displacement is needed on each end. With identical areas on either side of the piston, they can provide equal speeds and/or equal forces in either direction. Any doubleacting cylinder may be used as a single-acting unit by draining the inactive end to tank.

Lesson Summary
In this lesson we have dealt with the following topics: A. Basic Principles and Advantages of Hydraulic Control Systems: It is seen that force can be effectively multiplied by Hydraulic Systems due to Pascal’s Law. Further, there are several advantages of such systems with respect to motion control such as the ability for sudden stalling or reversal of motion under high loads. B. Hydraulic Fluids, Lines, Reservoirs, Filters and Seals : The functions of the fluid in the system is explained along with the accessories that carry it, such as lines and reservoirs. Other accessories such as filters and seals have also been presented briefly. C. Hydraulic Pumps and Accumulators: Various types of hydraulic pumps, namely, gear pimps, vane pumps and piston pumps have been considered and their principles of operation and construction explained. Two types of accumulators which act as temporary sources of fluids during transient high demand periods have also been presented. D. Hydraulic Motors and Cylinders: Factories have been classified into four major categories based on the product volumes and product variety. Similarly Automation Systems are also categorized into four types and their appropriateness for the various categories of factories explained.

Version 2 EE IIT, Kharagpur 15

Exercises
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. State Pascal's Law Name several advantages of a hydraulic system What makes petroleum oil suitable as a hydraulic fluid? What determines the speed of an actuator? How do you find the horsepower in a hydraulic system? Name three kinds of working lines and tell what each does Name four primary functions of the hydraulic fluid. Name four quality properties of a hydraulic fluid Name three functions of the reservoir? What are the basic characteristics of positive displacement pumps? How much oil does a vane pump rated for 5 gpm at 1200 rpm deliver at 1800 rpm? What tends to limit the pressure capability of a gear pump? What holds the vanes extended in a pump? How can displacement be varied in a axis piston pump? Name two functions of an accumulator.

Version 2 EE IIT, Kharagpur 16

Answers, Remarks and Hints to Points to Ponder Point to Ponder: 1
Can you give an analogy of the force amplification in hydraulic system from an electrical system? Ans: The electrical analog of force is voltage. Both are called across variables, while the electrical analog of flow rate is current, both which are called through variables. Note that the product of force and flow rate is power as is the product of voltage and current. Thus the analogy of force amplification is voltage amplication as can be achieved by transformers. A. Can you imagine what would happen, if the cylinder piston in Fig. 26.1 is stopped forcefully? Ans: If the cylinder is stopped, there cannot be any flow through the system. However, the prime mover to the pump would attempt to rotate the drive shaft and deliver fluid. Thus the operating pressure of the pump and load on the prime mover would tend to rise. Practically, this operating pressure would be contained by a relief valve which would open a low flow resistance path for the fluid to flow bypassing the cylinder (not shown in the Figure 26.1). Otherwise the load on the prime mover would be so high that it would stall. Thirdly, due to extremely high pressures fluid lines or pump may rupture.

Point to Ponder: 2
A. Consider two types of variable speed drives. In the first one an electric motor with a power electronic servo drive is directly coupled to the load through a mechanism. In the second one an electric motor with a constant speed drive drives the pump in a hydraulic system which provides the variable speed drive to the load. Which one of these two is more energy efficient? Ans: The first one is likely to be more efficient. This is because the overall efficiency of both the systems would include the efficiency of the motor and the efficiency of the final mechanism that connects the load with the actuator, such a gear or a ball screw. However, the hydraulic system would further involve the efficiency of the pump and cylinder as well as that of other speed control equipment such control valves. For the first system this would involve only the efficiency of the power electronic converter, which is likely to be higher. Thus the lesson is that hydraulic systems are not used for their energy efficiency, but rather for their small size, high power handling capacity and ease of control under high loads. B. Why is stalling an electric motor is likely to cause damage? What can be done to prevent it? Ans: Stalling an electric motor reduces the back emf in the motor to zero. Therefore very high current flows in the motor causing thermal damage. To prevent such damages, current control techniqies are applied in all motor drives which sense the current and reduce the motor terminal voltage whenever the current exceeds its limit. In other cases, where such rise Version 2 EE IIT, Kharagpur 17

of current is considered to be due to fault, over current trip mechanisms are employed that switch off supply to the motor.

Point to Ponder: 3
A. What would happen if orifices of valves are blocked by, say, a metal chip in the hydraulic oil? Ans: Immediately the pressure difference across the hydraulic cylinder, which moves the cylinder against load, would be neutralized. Thus the load motion would stop. At the same time the pressure difference across the jammed orifice would rise. Sometime this resulting force can dislodge or shear the chip that causes the jam.

Point to Ponder: 4
A. Why do gear pumps usually operate at comparatively low pressures? Ans: The load imposed by the drive shaft depends on the operating pressure. By construction, this load is unbalanced in the gear pump and therefore, considerable side loading on the drive shaft exists. To limit this loading, operating pressures have to be kept low. Note that due to the symmetry of the inlet and out let ports such forces do not arise in balanced vane pumps.

Version 2 EE IIT, Kharagpur 18

Module 6
Actuators
Version 2 EE IIT, Kharagpur 1

Lesson 27
Directional Control Valves, Switches and Gauges
Version 2 EE IIT, Kharagpur 2

Lesson Objectives
After learning the lesson students should be able to • • Describe the major types of direction control valves, their construction, operation and symbol Describe the major types of pressure relief and flow control valves, their construction, operation and symbol systems

• Describe pressure switches, as well as pressure and flow gauges used in hydraulic

Introduction
There are two basic types of hydraulic valves. Infinite position valves can take any position between fully open and fully closed. Servo valves and Proportional valves are in this category and are discussed in a separate lesson. Finite position valves can only assume certain fixed positions. In these positions the various inlet and outlet ports are either fully open or fully closed. However, depending on the position of the valve, particular inlet ports get connected to particular other outlet ports. Therefore flows in certain directions are established, while those in other directions are stopped. Since it is basically the directions of the flows that are controlled and not the magnitudes, these are called directional control valves. Directional valves can be characterized depending on the number of ports, the number of directions of flow that can be established, number of positions of the valve etc. They are mainly classified in terms of the number of flow directions, such as one-way, two way or four-way valves. These are described below. Directional valves are often operated in selected modes using hydraulic pressure from remote locations. Such mechanisms are known as pilots. Thus a valve that may be blocking the flow in a certain direction in absence of pilot pressure, may be allowing flow, when pilot pressure is applied. This enables one to build greater flexibility in the automation of system operation.

Check Valve
In its simplest form, a check valve is a one-way directional valve. It permits free flow in one direction and blocks flow in the other. As such is analogous to the electronic diode.

Fig. 27.1 The Check Valve Symbol The simple ball-and-seat symbol shown in Fig. 27.1 is used universally for denoting check valves although it is different from the way the other valves are denoted. The direction of the arrow shows the direction for free flow. Version 2 EE IIT, Kharagpur 3

In its simplest form of construction, a check valve is realized as an in-line ball and spring as shown in Fig.27.2. Seat Ball (or Poppet)

In

Out

Free Flow Allowed As Ball Unseats Fig. 27.2a A check valve permits flow in one direction Pressure from the left moves the ball from its seat so to permit unobstructed flow. Pressure from right pushes the ball tight on to the seat, and flow is blocked shown in Fig.27.2b. In some valves a poppet is used in place of the ball. In some other construction, the valve inlet and outlet ports are made at right angles.

Flow Blocked As Valve Seats Fig. 27.2b A Check Valve blocks flow in the reverse direction

Pilot – operated Check Valves
Pilot-operated check valves are designed to permit free flow in one direction and to block return flow, unless pilot pressure is applied. However, under pilot pressure, flow is permitted in both directions. They are used in hydraulic presses as prefill valves – to permit the main ram to fill by gravity during the “fast approach” part of the stroke. They also are used to support vertical pistons which otherwise might drift downward due to leakage past the directional valve spool. The construction of a pilot operated check valve is shown in Fig. 27.3. With no pilot pressure, the valve functions as a normal check valve. Flow to bottom is permitted but the reverse is blocked. If pilot pressure is applied, the valve is open at all times, and flow is allowed freely in both directions as shown in Fig 27.3b.

Version 2 EE IIT, Kharagpur 4

No Pilot Pressure

Drain

Free Flow Fig. 27.3a Unidirectional flow without pilot pressure No Pilot Pressure

Drain

No Flow Fig. 27.3b Unidirectional flow without pilot pressure

Version 2 EE IIT, Kharagpur 5

Pilot Pressure

Drain

Reverse Free Flow Fig. 27.3c Reverse flow with pilot pressure The check valve poppet has the pilot piston attached to the poppet stem. A light spring holds the poppet seated in a no-flow condition by pushing against the pilot piston. A separate drain port is provided to prevent oil from creating a pressure buildup on the underside of the piston. Reverse flow can occur only when a pressure that can overcome the pressure in the outlet chamber is applied. Possible application of the valve can be to permit free flow to the accumulator, while blocking flow out of it. If the pilot is actuated the accumulator can discharge if the pressure at the inlet port is lower than the accumulator pressure.

Point to Ponder: 1
A. Can you draw an inlet pressure flow characteristics of the ball type check valve? B. What should be the nature of the spring used in the check valves, hard or soft?

Two-Way and Four-Way Valves
Two-way and four-way valves direct inlet flow from the pump to the system through either of two outlet ports. They typically have a pump port, usually denoted as P, a tank port denoted T, and ports denoted A, B, which are connected to the system. These are finite position valves. In each position the ports P and T are connected to the ports A and B in such a fashion that a particular direction of flow is established. Finite position valves generally have four ports. Finite position valve symbols are constructed from squares. Typical applications include control of lift tables, press push backs, or other low pressure fluid devices. These valves can be classified either as rotary or as spool valves, depending on the mechanism that creates the connections in the various positions of the valve. These are conceptually

Version 2 EE IIT, Kharagpur 6

identical, only while the motion of the moving parts of the valve is rotational in the former and linear in the other. Therefore, in the sequel, mainly spool valves are discussed.

Rotary Valve
A rotary valve consists simply of a rotor closely fitted in a valve body. Passages in the rotor connect or block the ports in the valve body to provide the desired flow paths. They are used principally as pilot valves to control other valves.

Spool Type Valve
In the spool type directional valve a cylindrical spool moves back and forth in a machined bore in the valve body. The port connections in the body are interconnected through annular grooves in the spool or blocked by the raised portions of the spool called lands. Changes in valve operation are achieved by utilizing spools with different land patterns, with the same valve body.

Two way valve
Fig 27.4a represents, symbolically, a two way valve where in one position the port P is connected to the load B while port A and port T are plugged. In the other position, the port P is connected to the load A while port B and T remain plugged. The valve moves its position from port A to port B and vice versa when force is applied to the spool, the port A is selected with the spool to the right and the port B with the spool to the left. Tank only serves to drain leakage from the valve. A B

P

T

Fig 27.4a Symbol for a two-way valve A B

T

P

T

Fig 27.4b A two-way valve with the spool to the left Version 2 EE IIT, Kharagpur 7

A

B

T

P

T

Fig 27.4c A two-way valve with the spool to the right

Four way valve
In two-way valves the pump port is connected to ports A and B in the two positions and the tank port serves only to drain leakage from within the valve. Thus the return flow does not take place through the valve. In the four-way valve the ports P and T are connected to the ports A and B, respectively, in one position, and to B and A respectively, in the other. Thus, both the flow from the pump and the return flow to the reservoir are directed through the valve. Most of these valves are of the spool type. They are available both in two-position or threeposition versions. The three-position valve has a center or neutral position. Methods of actuation include manually operated levers, cams and linkages, solenoids, hydraulic or pneumatic pilot pressure etc. A B

P T Fig 27.5a Symbol for a four-way valve A B A B

P

T

T

P

Fig 27.5b A four-way valve with the spool to the left

Fig 27.5c A four-way valve with the spool to the left

Version 2 EE IIT, Kharagpur 8

Spool Center Conditions
A B

P

T

Open Center, all ports connected

A B

P

T

B Pump (P) and “B” Closed-“A” Open to Tank (T)

A B

P

T

T A B Pump (P) and “B” Closed-“A” Open to Tank (T)

A B

P

T Pump (P) Closed-“A” & “B” Open to Tank (T)

A B

P

T

B T A “B” Closed-Pump (P) Open to Tank (T) Thru “A”

A B

P

T

A

B

Tandem Fig. 27.6 Various center Configurations and their Symbols

Version 2 EE IIT, Kharagpur 9

Three-position valve spools provide identical flow patterns in the shifted positions but available with a variety of port connections configurations in the neutral position as illustrated in Figure 27.6. The open-center spool interconnects all ports and the pump delivery can flow to tank at low pressure. The closed center spool, on the other hand, has all ports blocked. This allows the pump delivery to be diverted to other parts of the circuit. Otherwise, it is to be diverted to tank through a relief valve. Other center conditions permit various other configurations blocking and opening of ports. Spools may be held in the center positions by centering springs, by spring-loaded detents or by oil under pressure. The terms “spring-centered” and “spring-offset” refer to the use of springs returning valve spools to their normal position. A spring centered valve is returned to the center position by spring force whenever the actuating effort is released. A spring-offset valve is a two position valve returned to one extreme position by a spring whenever the actuating effort is released. It is shifted to the other position by one of the actuating methods mentioned above. A valve without a spring has to be actuated throughout the cycle by an external actuation, without which, it may “float” unless retained by detents or friction pads.

Operating Controls
As mentioned above, spool valves can be actuated in a number of ways. A typical manuallyoperated four-way valve is shown with its graphical symbol in Figure 27.7a; a mechanicallyoperated valve in Figure 27.7b. Figure 27.7c illustrates a spool type four-way valve that is shifted by air pressure against a piston at either end of the valve spool. A common method of actuating a small spool valve is with a solenoid as shown in Fig. 27.7d. Note that the appropriate controlling symbol is added with the basic valve symbol. A B

P

T

Fig. 27.7a A solenoid operated three-position four-way directional valve symbol In larger valves, the force required to shift the spool is more than what can be obtained practically from a solenoid. Large directional valves are, therefore, actuated by pilot pressure (Fig. 27.7b). The pilot pressure, in turn, can be directed from a hydraulic pressure source by a smaller four-way valve called a pilot valve, which may be actuated by a solenoid. A B

P

T

Fig. 27.7b A solenoid operated three-position four-way directional valve symbol Version 2 EE IIT, Kharagpur 10

Point to Ponder: 2
A. In Fig. 27.7b, what is the position of the valve when there is no pressure at both the pilot ports? B. Is it possible to connect directional control valves in series and/or parallel? Can anything be achieved thereby? Are there conditions in which this is possible?

Relief Valves
Relief valves are used for regulation of pressure in hydraulic systems for protection of equipment and personnel. The simplest relief valve is the spring and plug arrangement of Fig. 27.8a. The spring keeps the valve shut until a pressure set by an adjustable spring tension is reached which pushes the spring up to relief the pressure by connecting the inlet to the drain. Adjust Spring Preloading

Plug

Drain

Inlet Fig 27.8a Relief valve Fig 27.8b is a symbol for a relief valve. The arrow on the spring shows adjustable tension. Pilot operated relief valves are shown with a dotted line and without a spring symbol.

Pilot Fig 27.8b Relief valve symbols Infinite position valves (example relief valves) have a single arrow. The arrow shows flow direction and are generally drawn in non operated condition.

Version 2 EE IIT, Kharagpur 11

Pressure Switches
Pressure switches (Fig. 27.9) are used to make or break (open or close) electrical circuits at selected pressures to actuate solenoid operated valves or other devices used in the system. The operating principle of a pressure switch is shown in Figure 27.9. This design contains two separate electrical switches. Each is operated by a push rod which bears against a plunger whose position is controlled by hydraulic and spring forces. The pressure at which the switches operate is selected by turning the adjusting screw to increase or lessen the spring force. It should be noted that in this design the switches are actuated by the springs when the unit is assembled. Thus the normally open contacts are closed and vice versa. When the preset pressure is reached the plunger will compress the spring and allow the push rods to move down causing the snap action switches to revert to their normal condition. By using both switches in conjunction with an electrical relay, system pressures may be maintained within widely variable high and low ranges.

At high pressure setting, this plunger moves down to open back switch completing circuit shown by dashed line.

Instruments
Flow rate, pressure and temperature measurements are required in evaluating the performance of hydraulic components. All three can be helpful too in setting up or trouble-shooting a hydraulic system. Due to the difficulty in installing a flow meter in the circuit, flow measurements are Version 2 EE IIT, Kharagpur 12

often determined by timing the travel or rotation of an actuator. Pressure and temperature are determined in the usual manner by means of gauges and thermometers.

Pressure Gauges
Pressure gauges are needed for adjusting pressure control valves to required values and for determining the forces being exerted by a cylinder or the torque of a hydraulic motor. A typical pressure gauge used is a Bourdon tube. In the Bourdon tube gauge (Fig. 27.10), a sealed tube is formed in an arc. When pressure is applied is applied at the port opening, the tube tends to straighten. This actuates linkage to the pointer gear and moves the pointer to indicate the pressure on a dial. Most pressure gauges read zero at atmospheric pressure and are calibrated in pounds per square inch, ignoring atmospheric pressure throughout their range. Pump inlet conditions are often less than atmospheric pressure. They would have to be measured as absolute pressure, sometimes referred to as psia, but more often calibrated in inches of mercury with 30 inches of mercury considered a perfect vacuum. A vacuum gauge calibrated in inches of mercury is shown in figure 12-11. Tube tends to straighten under pressure causing pointer to rotate. 70 80 90 0 100

40 30 20 10 Bourdon tube

50

60

Pressure inlet Fig. 27.10 A C type Bourdon gauge

Flow Meters
Flow meters are coupled into the hydraulic piping to checkihe volumetric efficiency of a pump and determining leakage paths within the circuit. A typical flow meter used is a rotameter, which consists of a weight in a calibrated vertical tube. Oil is pumped into the bottom and out the top and raises the weight to a height proportional to the flow. For more accurate measurement, a fluid motor of known displacement can be used to drive a tachometer. More sophisticated measuring devices are turbine type flow meters which generate an electrical impulse as they rotate and pressure sensing transducers which may be located at strategic points Version 2 EE IIT, Kharagpur 13

within the system where they send out electrical signals proportional to the pressures encountered. These signals can be calibrated and observed on an oscilloscope or other readout devices.

Point to Ponder: 3
A. Consider the type 4 and type 6 centre conditions of the 3-position 4 way valve shown above. Imagine that a double acting hydraulic cylinder is connected across the ports A and B. Comment on the possible motion of the load when the valve is at the center. B. Can you justify, why the bourdon tube is perhaps the most suitable of the pressure sensor sin this application?

Lesson Summary
In this lesson we have dealt with the following topics A. Basic types of direction control valves, such as one-way, two-way and four-way valves. B. Introduced the concept of pilot operation of valves in the context of check valves and relief valves. C. Described various center conditions as well as operating controls for fourway valves. D. Described typical types of instruments and switches used in hydraulic circuits.

Version 2 EE IIT, Kharagpur 14

Answers, Remarks and Hints to Points to Ponder Point to Ponder: 1
A. Can you draw an inlet pressure flow characteristics of the ball type check valve? Ans: Let a positive pressure drop correspond to the forward flow direction. Then, because of spring force, a minimum positive pressure drop would be needed to establish full flow. Similarly, below a minimum pressure drop, no flow would take place. For negative pressure drop, the flow is zero. B. What should be the nature of the spring used in the check valves, hard or soft? Ans: With a hard spring, because of spring force, the minimum positive pressure drop needed to establish full flow would be high. Therefore the spring should be soft, only ensuring that when the pressure drop across the valve is very small, the flow is cut off.

Point to Ponder: 2
A. In Fig. 27.7b, what is the position of the valve when there is no pressure at both the pilot ports? Ans: The valve shown is a two-position four-way valve. The spring symbol indicates that if no pilot pressure exists at either port, the position is determined by the spring. Since the spring is shown at the left, the left box shows the position of the valve when pilot pressures at the ports are zero. B. Is it possible to connect directional control valves in series and/or parallel? Can anything be achieved thereby? Are there conditions in which this is possible? Ans: The valves can easily be connected in parallel, provided the pump is capable of delivering the required flow. In case they are connected in series, the tank port of the first valve is connected to the pump port of the other. For two position valves this is possible. However for three position valves, for independent operation of the two loads, the valve should be open center type, that is the pump and tank ports should be connected. If these ports remain plugged then independent operation of loads would not be possible.

Point to Ponder: 3
A. Consider the type 4 and type 6 centre conditions of the 3-position 4 way valve shown above. Imagine that a double acting hydraulic cylinder is connected across the ports A and B. Comment on the possible motion of the load when the valve is at the center. Ans: For the type 4 valve, the load is locked, since both the ports A and B are plugged. For the type 6 valve, the load is floating, since both the ports A and B are connected and are at tank pressure. Therefore with any pressure applied on the load, the cylinder can shift without resistance. Version 2 EE IIT, Kharagpur 15

B. Can you justify, why the bourdon tube is perhaps the most suitable of the pressure sensor sin this application? Ans: For hydraulic circuits, continuous pressure monitoring is not needed. Generally intermittent manual monitoring is needed for maintenance and trouble shooting. Therefore gauges such as diaphragm gauges, which can operate at similar ranges of pressure and can provide continuous electrical signals is a more expensive alternative. Bourdon gauges are inexpensive, rugged and fairly accurate for the application.

Version 2 EE IIT, Kharagpur 16

Module 6
Actuators
Version 2 EE IIT, Kharagpur 1

Lesson 28
Industrial Hydraulic Circuits
Version 2 EE IIT, Kharagpur 2

Lesson Objectives
After learning the lesson students should be able to • • • Describe typical industrial actuation problems Interpret hydraulic system symbols and circuit diagrams Describe techniques for energy saving in hydraulic systems

Introduction
Typical hydraulic circuits for control of industrial machinery are described in this lesson. Graphical hydraulic circuit diagrams incorporating component symbols are used to explain the operation of the circuits.

Case Study I: Unloading System for Energy Saving
An “unloading” system is used to divert pump flow to a tank during part of the operational cycle to reduce power demand. This is done to avoid wasting power idle periods. For example, it is often desirable to combine the delivery of two pumps to achieve higher flow rates for higher speed while a cylinder is advancing at low pressure. However, there may be considerable portions of the cycle, such as when the cylinder is moving a heavy load, when the high speed is no longer required, or cannot be sustained by the prime mover. Therefore, one of the two pumps is to be unloaded resulting in a reduction of speed and consequently, power. The components of this system are: A, B: Hydraulic pumps, C, E: Pilot operated Spring loaded Relief valves, D: Check valve

Mode 1: Both Pumps Loaded
In Figure 28.1 below, when both pumps are delivering, oil from the pump A passes through the unloading valve C and the check valve D to combine with the pump B output. This continues so long as system pressure is lower than the setting of the unloading valve C.

Version 2 EE IIT, Kharagpur 3

D

E C

A

M

B

Fig. 28.1 Unloading Circuit

Mode 2: One pump unloaded
In Fig. 28.1, when system pressure exceeds the setting of the unloading valve C, it makes pump A to discharge to the tank at little pressure. Although the system pressure, supplied by pump B, is high, the check valve prevents flow from B through the unloading valve. Thus only pump B now drives the load at its own delivery rate. Thus the load motion becomes slower but the power demand on the motor M also reduces. If the system pressure goes higher, say because load motion stops, pump B discharges when its relief valve settings would be exceeded.

Points to Ponder: 1
A. Can you imagine what would happen, if the check valve was not present? B. How would you modify the system if you wanted to unload pump B instead of pump A?

Case Study II: Selection of System Operating Pressure
The circuit shown in Figures 28.2-28.4 allow selection of operating pressure limits in a hydraulic system from three options, namely, two maximum pressures, plus venting. First note the components, namely, A: Reservoir with Filter, B: Hydraulic Pump, C, E: Pilot Relief Valve, D: Solenoid activated Four-way Directional valve.

Venting Mode
In Figure 28.2, both solenoids a and b of the directional valve D are de-energized. The opencenter spool is centered by the valve springs, and the vent port on the relief valve is opened to tank. Therefore, the pump flow opens to tank at a very low pressure. Version 2 EE IIT, Kharagpur 4

E D A a P C To System T B b

B

M
A

Fig. 28.2 Venting Mode

Intermediate Maximum Operating Pressure
In Figure 28.3, the left-hand solenoid a of the directional valve is energized. The valve spool is shifted to the leftmost position and connects the relief valve vent port to the remote control valve. Pump flow is now diverted to tank when the pressure setting of the remote valve E is reached. E D A a P C To System T B b

B

M
A

Fig. 28.3 Operating Mode with Intermediate Maximum Operating Pressure Version 2 EE IIT, Kharagpur 5

High Maximum Pressure
In Figure 28.4, the right solenoid b of the directional valve is energized. The spool now shifts right to connect the relief valve vent port to a plugged port in the directional valve. The relief valve C now functions at the setting of its integral pilot stage. E D A a P C To System T B b

B

M
A

Fig. 28.4 Operating Mode with High Maximum Operating Pressure

Points to Ponder: 2
A. Why are lines connecting C to D and D to E marked in dashed lines? B. Can you briefly describe a scheme to automate the above system such that whenever, in the intermediate pressure mode, pressure setting is exceeded, the system would automatically switch to the low pressure mode?

Case Study III: Reciprocating Cylinder with Automatic Venting at End of Cycle
A reciprocating cylinder drive is a very common hydraulic system. In systems where it is not necessary to hold pressure at the end of a cycle, it is desirable to unload the pump by automatically venting the relief valve, to save energy. Figures 28.5-28.8 show such a system. The system components are : A : Reservoir with Filter, B : Hydraulic pump, C, E : Check valve, D : Pilot operated relief valve, F : Two-position electro-hydraulic pilot operated Four-way Directional valve, G : Cam operated pilot valve, H : Double acting Single rod Cylinder, I : Limit Switch.

Version 2 EE IIT, Kharagpur 6

Extension Stroke
Consider the beginning of the machine cycle when the solenoid of the spring offset directional valve F is energized. Pump output is connected to the cap end of the cylinder. The vent line drawn from the directional valve output connected to the cap end of the cylinder is blocked at the cam-operated pilot valve G. Thus, vent port of the relief valve D is blocked, and the cylinder moves under full pump pressure applied to the cap end. H

G F A E D b P T C B

B

Fig. 28.5 Extension Stroke

Retraction Stroke
At the extreme end of the extension stroke, the limit switch is made on by the cylinder rod to break the solenoid circuit for the directional valve F. The directional valve now shifts to its right position and the pump gets connected to the rod end of the cylinder which now retracts. Note that the relief valve vent connection is still blocked.

Automatic Venting at End of Retraction Stroke
At the extreme end of the retraction stroke, the cam on the cylinder is operated by the rod

Version 2 EE IIT, Kharagpur 7

H

G A E D P B T DR C F

B

A Fig. 28.6 Retraction Stroke to shift valve G. The relief valve vent port is thus connected, through E and G, to the line from the cap end of the cylinder, and to tank through the F and the inline check valve C. This vents the relief valve D and unloads the pump.

Push Button Start of Cycle
If another cycle of reciprocating motion is desired, a start button connected to the solenoid circuit is depressed to energize the solenoid, and, in turn, the directional valve shifts to direct pump output into the cap end of the cylinder. This causes the check valve in the vent line to close. Pressure again builds up and the cylinder starts extending. This releases the cam, which, under spring action, shifts and the vent port of E is again blocked at G. Thus the cycle repeats.

Points to Ponder: 3
A. How does the solenoid get energized if the limit switch is made? B. Is the speed of the cylinder going to be equal during extension and retraction? If not, then what decides the speeds?

Version 2 EE IIT, Kharagpur 8

H G 1 2 A E D P B T DR C F

B

A Fig. 28.7 Automatic Venting at End of Retraction Stroke

H G 1 2 F A E D P B T DR C

B

A Fig. 28.8 Push Button Start of Cycle

Version 2 EE IIT, Kharagpur 9

Case Study IV: Regenerative Reciprocating Circuit
Conventional reciprocating circuits use a four-way directional valve connected directly to a cylinder. In a regenerative reciprocating circuit, oil from the rod end of the cylinder is directed into the cap end to increase speed, without requiring to increase pump flow. Such a circuit is shown below in Figures 28.9-28.10. The circuit components are : A : Hydraulic Pump, B : Relief valve, C : Four-way two position solenoid operated valve, D : Double-acting Single-rod Cylinder. The operation of the regenerative circuit is shown in Figures 28.9-28.10.

Regenerative Advance
In Figure 28.9, the “B” port on the directional valve C, which conventionally connects to the cylinder, is plugged and the rod end of the cylinder is connected directly to the pressure line. With the valve shifted to the left most position, the “P” port is connect to the cap end of the cylinder. If the ratio of cap end area to rod end annular area in the cylinder is 2:1, the pressure being the same at both end, the force at the cap end is double that at the rod end. There is therefore a net force on the cylinder to move the load. Similarly, at any speed of the cylinder, the flow into the cap end would be double that of the rod end. However, in this connection, the flow out of the rod end joins pump delivery to increase the cylinder speed. Thus only half of the flow into the cap end is actually supplied by the pump. However, the pressure during advance will be double the pressure required for a conventional arrangement for the same force requirement. This is because the same pressure in the rod end, effective over half the cap end area, opposes the cylinder’s advance. In the reverse condition shown in Figure 28.10, flow from the pump directly enters the rod end of the cylinder through two parallel paths, one through the directional valve and the other directly. Exhaust flow from the cap end returns to the tank conventionally through the directional valve. Note that, in contrast to the conventional case, the force on the cylinder as well as the pump flow remains unchanged during extension and retraction. Thus, the speed of the piston during both advancement and retraction remain same.

A B a P T b

Fig. 28.9 Regenerative Extension Stroke

A B a P T b

Fig. 28.10 Regenerative Retraction Stroke Version 2 EE IIT, Kharagpur 10

Points to Ponder: 4
A. Explain all parts of the symbol of the directional valve C in Figures 28.9-28.10. B. Compare, point by point a regenerative reciprocation circuit with a conventional one.

Case Study V: Sequencing Circuits
In many applications, it is necessary to perform operations in a definite order. Following is one of several such circuits. The components of the system are as follows. A : Reservoir and Filter ; B : Hydraulic Pump ; C : ; Relief valve : D ; F1, F2, G : Relief valve with integral check valve ; H, J : Cylinders ; I : Check Valve The sequence of operation realized by the circuit shown in Figure 28.11-28.14 is: Step A – Extend Cylinder H Step B – Extend Cylinder J while holding pressure on Cylinder H Step C – Retract Cylinder J Step D – Retract Cylinder H D F1 E B P C T A B G H J

F2

Fig. 28.11 Cylinder H Extending

Version 2 EE IIT, Kharagpur 11

Step A
Pressing a pushbutton would start the cycle and shift the directional valve E to the position shown in Fig 28.11. At first the fluid flows through the integral check valve in G into the cap end of H and returns freely through the check valve in F2. The pump pressure is low during this period, only to the extent of pushing the load on H.

Step B
Once H reaches its rod end, the pressure builds up and now the flow develops through F1 into the cap end of J and out through the rod end to go back directly to tank through F2, E and C. Note that a pressure equal to the setting of the valve F1 is maintained on H. When J is fully extended, pressure increases further and is limited by the setting of D, providing overload protection to B. D F1 E B P C T A B G H J

F2

Fig. 28.12 Cylinder J Extending

Step C
Similarly, when the other solenoid of E is energized, the directional valve shifts to the other position, as shown in Fig. 28.13. Now, pump delivery is directed through D, E and F2, into the rod end of J. As before, the flow out of the cap end of J flows to tank through F1, E and C. Step C is illustrated in Fig. 28.13. Version 2 EE IIT, Kharagpur 12

Step D
On completion of Step C, the pressure increases again, and the flow is directed through F2 to the rod end of H and out through the cap end to flow into the tank through the valve G at its pressure setting and then freely to tank through F1, E and C. Note that F2 maintains a pressure equal to the setting of H at the rod end of J during the retraction of H. Note further that, while H is retracting, a back pressure is provided to it by G, to prevent rapid falling of the load during lowering, under gravity. In the above circuit, sequencing is achieved by grading the pressure settings of the relief valves. Note that sequencing can also be achieved electronically by PLC control of the solenoids of separate directional valves driving H and J. D E B J

F1

G

H

C

F2

Fig. 28.13 Cylinder J Retracting

Version 2 EE IIT, Kharagpur 13

D E B

F1

J

C G H

F2

Fig. 28.14 Cylinder H Retracting

Lesson Summary
In this lesson we have presented industrial hydraulic circuits for following applications: A. Unloading of pumps based on relief valve and check valve to avoid overloading. B. Selection of system operating pressure from two values based on settings of relief valves and their selection with a directional valve. C. Extension-retraction Circuit for one cylinder with automatic unloading of pump at the end of cycle D. Extension-retraction Circuit for one cylinder with regenerative feedback E. Sequenced extension-retraction circuit for two cylinders.

Version 2 EE IIT, Kharagpur 14

Answers, Remarks and Hints to Points to Ponder Points to Ponder: 1
A. Can you imagine what would happen, if the check valve was not present? Ans: Then, if the setting of the relief valve C is reached, both pumps would have unloaded. B. How would you modify the system if you wanted to unload pump B instead of pump A? Ans: Then, a check valve is to be placed between pump B and the system, after the relief valve D. The relief valve after C would be removed. Finally, the setting of D would have to be set lower than that of C.

Points to Ponder: 2
A. Why are lines connecting C to D and D to E marked in dashed lines ? Ans: Because these are pilot lines for the relief valve. B. Can you briefly describe a scheme to automate the above system such that whenever, in the intermediate pressure mode, pressure setting is exceeded, the system would automatically switch to the low pressure mode? Ans: One of the ways is to devise an automated logic for exciting the solenoids of the directional valve. The above feature can be implemented by using a pressure switch which would be made if the pressure setting is exceeded. This pressure switcvh contacts can then be incorporated into a relay circuit or a PLC program that would drive the solenoid of the directional valve corresponding to high pressure.

Points to Ponder: 3
A. How does the solenoid get energized if the limit switch is made? Ans: The limit switch contacts can be incorporated into a relay circuit or a PLC program that would drive the solenoid of the directional valve. B. Is the speed of the cylinder going to be equal during extension and retraction? If not, then what decides the speeds? Ans: The speed of the cylinder is going to be different during extension and retraction. Note that the cross-sectional areas of the two sides of the piston are different because of the single-rod configuration. During extension, the pump pressure acts on the cap end of the cap end of the cylinder which has higher area than the rod end. Therefore, the net force acting on the piston during extension is more than that during retraction. The speeds reached during these strokes depend on the load that exists during these

Version 2 EE IIT, Kharagpur 15

strokes. Thus, to summarize, the speeds depend on the loads as well as the areas of the rod and the cap ends f the cylinder.

Points to Ponder: 4
A. Explain all parts of the symbol of the directional valve C in Figures 28.9-28.10. Ans: The directional valve has two positions and is a four-way valve with the pump port and the tank port on one side and the two load side ports A and B on the other. The valve is not spring loaded and therefore requires two solenoids for moving it to the two positions. The solenoid moves a hydraulic pilot valve spool, which, in turn, moves the main valve spool. The valve is called detented because it has a mechanism that locks the piston into the position it moves into and holds it there. In the circuit in which the valve is used, in the left position, one port of the valve is plugged, as marked by a cross. B. Compare, point by point a regenerative reciprocation circuit with a conventional one. Ans: In a regenerative circuit higher cylinder speed can be reached with a lower pump flow rate. However, for a given force requireent to move the load, higher pup pressures are needed, since there exists a back pressure at the rod end of the piston, unlike in a conventional reciprocation circuit. In the retraction stroke, however, regeneration is not possible. In both cases, regenerative and conventional, cylinder speeds during retraction and extension are different in general. Or the regenerative case, however, they may be made equal with area ration of 2:1 of the cap and the rod ends of the piston.

Version 2 EE IIT, Kharagpur 16

Module 6
Actuators
Version 2 EE IIT, Kharagpur 1

Lesson 29
Pneumatic Control Components
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
At the end of this lesson, the student should be able to • • • • • • • Explain with a sketch the principle of operation of a flapper nozzle amplifier. Derive the approximate relationship between the output pressure and displacement for a flapper nozzle amplifier. Justify the use of air relay in conjunction with a flapper nozzle amplifier. Explain the advantage of using closed loop configuration of flapper nozzle amplifier. Sketch and explain the operation of a flapper nozzle amplifier in closed loop. Explain the limitation of a direct acting type valve positioner. Explain the principle of operation of a feedback type valve positioner.

Introduction
A number of pneumatic components are present in a process control scheme. In earlier days, the complete control system was built up on these components; with the advent of electronics many of them are now replaced by electronic components. Still then, the importance of the pneumatic components cannot be underestimated. Many of the industrial actuators used in steel and automobile industries nowadays are pneumatic. The major advantages of using pneumatic systems are (i) they are intrinsically safe and can be used in hazardous atmospheres, (ii) cheap compared to hydraulic systems (air costs nothing) and (iii) a pneumatic actuator can generate more torque (force) to its own weight and thus have a better torque-weight ratio compared to an electrical actuator. However pneumatic components are slow in response. In this lesson we will discuss different pneumatic components used in process control.

Flapper nozzle amplifier
A pneumatic control system operates with air. The signal is transmitted in form of variable air pressure (often in the range 3-15 psi, i.e. 0.2 to 1.0 bar) that initiates the control action. One of the basic building blocks of a pneumatic control system is the flapper nozzle amplifier. It converts very small displacement signal (in order of microns) to variation of air pressure. The basic construction of a flapper nozzle amplifier is shown in Fig.1. Constant air pressure (20psi) is supplied to one end of the pipeline. There is an orifice at this end. At the other end of the pipe there is a nozzle and a flapper. The gap between the nozzle and the flapper is set by the input signal. As the flapper moves closer to the nozzle, there will be less airflow through the nozzle and the air pressure inside the pipe will increase. On the other hand, if the flapper moves further away from the nozzle, the air pressure decreases. At the extreme, if the nozzle is open (flapper is far off), the output pressure will be equal to the atmospheric pressure. If the nozzle is blockes, the output pressure will be equal to the supply pressure. A pressure measuring device in the pipeline can effectively show the pressure variation. The characteristics is inverse and the pressure decreases with the increase in distance. Typical characteristics of a flapper nozzle amplifier is shown in Fig.2. The orifice and nozzle diameter are very small. Typical value of the orifice diameter is 0.01 inch (0.25 mm) and the nozzle diameter 0.025 inch (0.6 mm). Typical change in pressure is 1.0 psi (66 mbar) for a change in displacement of 0.0001 inch (2.5 micron). Version 2 EE IIT, Kharagpur 3

There is an approximate linear range in 3-15 psi, of the characteristics of the amplifier, that is the normal operating range. Nozzle Air Supply xi Output pressure po (3-15psi) Flapper

po

Fig. 1 Flapper nozzle amplifier po(psi) 20 15 Approximate linear range

10

3 0 5 10

xi (mil)

Fig. 2 Characteristics of a flapper nozzle amplifier.

Performance Analysis
The performance analysis of the flapper nozzle amplifier can be carried out in two ways: neglecting the compressibility of air and taking compressibility of air into account. For the sake of simplicity, we shall neglect the compressibility in this section and carry out the simplified analysis. The mass flow rate through the orifice can be expressed as: Cd π d s2 (1) 2 ρ ( ps − po ) Gs = 4

Version 2 EE IIT, Kharagpur 4

where, Cd is the discharge coefficient of the orifice, ds is the inside diameter of the orifice, ρ is the density of air, ps is the supply pressure and po is the pressure inside the pipe. The above expression comes directly from the Bernoulli’s equation, considering that the area of the orifice is much smaller than the area of the pipe. For finding out the flow through the nozzle, the flow area is taken as the peripheral area of a cylinder of diameter dn (nozzle diameter) and length xi (distance between the flapper and the nozzle). That means that if we imagine a cylinder of diameter dn and length xi, the air is going out of the nozzle to the atmosphere in the radial direction and the area of the orifice thus formed will be surface area of the cylinder. Noting that the air pressure outside the cylinder surface is ambient pressure (pamb), similar to (1), we can write the expression for the mass flow rate through the nozzle as:

Gn = Cd π d n xi 2 ρ ( po − pamb )

(2)

We have assumed air to be incompressible. The discharge coefficient is also assumed to be the same for both the orifice and the nozzle. So at steady state,

Gs = Gn , and pamb = 0 .
Equating (1) and (2) and simplifying, one can obtain:

d s4 ( ps − po ) = d n2 xi2 po 16
or,

po = ps

1 2 16d n 2 1 + 4 xi ds

(3)

Now denoting the normalized pressure pn = we can write,

po d , and the normalized displacement as xn = n xi , ps d s2

pn =

1 2 1 + 16 xn

(4)

The pn vs. xn characteristics is similar to that shown in Fig.2. The sensitivity can be obtained as:

dpn 1 = −32 xn 2 dxn (1 + 16 xn )2
For sensitivity to be maximum,
2 2 d 2 pn 32(1 + 16 xn ) 2 − 32 xn .2(1 + 16 xn ).32 xn =0=− 2 2 dxn (1 + 16 xn ) 4

(5)

Solving, one obtains the condition for maximum sensitivity as:
2 xn =

1 ; or xn ≈ 0.144 48

The maximum sensitivity, at xn = 0.144 is
Version 2 EE IIT, Kharagpur 5

dpn = −2.59 dxn and at this value of xn, 1 1 3 pn = = = = 0.75 2 1 + 16 xn 1 + 16. 1 4 48 If the supply pressure is 20 psi, the sensitivity is maximum when the output pressure p0 is around 15 psi. In order to avoid zero or very low sensitivity, the minimum workable pressure is chosen as 3 psi. Thus the working output pressure range of 3-15 psi is normally used for practical applications.

Air Relay
The major limitation of a flapper nozzle amplifier is its limited air handling capacity. The variation of air pressure obtained cannot be used for any useful application, unless the air handling capacity is increased. The situation can be compared with an operational amplifier in an electronic circuit. Though the operational amplifier is useful in amplifying small voltage signals, the output current delivered by the operational amplifier is limited and a power amplifier is used at the output stage in order to drive any device. An air relay serves the similar purpose as a power amplifier. It is used after the flapper nozzle amplifier to enhance the volume of air. The principle of operation of an air relay can be explained using the schematic diagram shown in Fig. 3.

Version 2 EE IIT, Kharagpur 6

Air Supply (ps)

double seated valve

pout Air vent y p2 Diaphragm p2

xi Fig. 3 Schematic diagram of an air relay

It can be seen from Fig.3 that the air relay is directly connected to the supply line (no orifice in between). The output pressure of the flapper nozzle amplifier (p2) is connected to the lower chamber of the air relay with a diaphragm on its top. The variation of the pressure p2 causes the movement (y) of the diaphragm. There is a double-seated valve fixed on the top of the diaphragm. When the nozzle pressure p2 increases due to decrees in xi, the diaphragm moves up, blocking the air vent line and forming a nozzle between the output pressure line and the supply air pressure line. So more air goes to the output line and the air pressure increases. When p2 decreases, the diaphragm moves downward, thus blocking the air supply line and connecting the output port to the vent. The air pressure will decrease.

Flapper Nozzle Amplifier with Feedback
Another problem of a flapper nozzle amplifier is its sensitivity variation. It can be easily seen from Eqn. (3) that the output pressure p0 is dependent on the supply pressure, orifice diameter and the nozzle diameter. Any variation of the supply pressure will affect its sensitivity. Moreover, accumulation of dirt at the nozzle or at the orifice will alter the sensitivity. As a result, some measure is needed to reduce this parameter dependence of the sensitivity. Use of feedback is an effective method for reducing the variation of the sensitivity. Flapper nozzle amplifiers are never used in open loop; it is always used in closed loop (we can draw an analogy with operation Version 2 EE IIT, Kharagpur 7

amplifiers in this respect: operational amplifiers are always used in closed loop). A typical application of flapper nozzle application with feedback for measurement of pressure and converting the signal in terms of air pressure variation is shown in Fig. 4. The scheme is called a torque balance arrangement.
PS (20psi)

Air Relay Input pressure Feedback Bellows (Area AB2) a pi b po xi Input Bellows (Area AB1) zero spring Fulcrum Fig. 4 Flapper nozzle amplifier with feedback. Fo

po

The basic scheme shown here has two bellows, one measuring the unknown pressure (pi); the other, known as output bellows is connected to the output pressure line of the system. These two bellows are attached to the two ends of a link, pivoted at some intermediate position. The link towards the output bellows is extended and forms the flapper of the flapper-nozzle amplifier. The output of the flapper nozzle amplifier is connected to the air relay whose output is the output pressure (p0) of the system. A spring is also attached to the link as shown in Fig.4. One end of the spring is fixed and the other end is connected to the link. The fixed end of the spring can be adjusted so that the spring generates a variable upward force F0 to the link. This spring is used for zero adjustment, say, when we want that p0 = 3psi for pi= 0. Suppose initially the rigid link is at stable horizontal position. In that case the clockwise and anticlockwise torques on the beam would balance. Looking at Fig.4, Anticlockwise moment: TA = Pi AB1 a + F0b , and Clockwise moment: TC = p0 AB 2b Where AB1 and AB 2 are the areas of the two bellows, a and b are the corresponding lengths of the link segments. Thereby at balance: AB a F (6) p0 = 1 pi + 0 AB2 b AB2 Version 2 EE IIT, Kharagpur 8

When the input pressure increases, the left side of the link moves down, thus moving the flapper on the right hand side closer to the nozzle. This will increase the nozzle pressure and subsequently the pressure p0 at the outlet of the air relay. The bellows in the right hand side is connected to this output pressure line. Increase in this pressure will result in more downward force by the output bellows, thus moving the nozzle back to almost its original position. From the expression given in (6), it is apparent that the output pressure here is independent of the diameters of the orifice and nozzle, thus is not affected by the accumulation of dirt or sensitivity variation due to variation of the supply pressure. Moreover the sensitivity can be adjusted by varying the lengths a and b.

Electro-pneumatic Signal Converter
It has been mentioned earlier, that the controller used in process control is normally electronic and for actuation pneumatic actuator is the preferred. Thus there is a need for converting the electrical signal (often 4-20 mA) from the controller to pneumatic 3-15 psi signal. Such a scheme is shown in Fig.5. It is similar to that one shown in Fig.4, except there is an electromagnet and a permanent magnet on the left of the link. The current flowing through the electromagnet causes a force of repulsion between the electromagnet and the permanent magnet. An increase in current through the coil increases the repulsive force, thereby moving the link upward on the left hand side and decreasing the gap between the flapper and the nozzle. The feedback action causes the increase in the output pressure and brings back the link in its equilibrium position.

4-20 mA

Air Relay Output (3-15psi)

Permanent magnet

PS (20psi) Fig. 5 Electro-pneumatic Signal Converter

Version 2 EE IIT, Kharagpur 9

Pneumatic Valve Positioner
Pneumatic valve positioner is another important component used in process control. The control valve should be moved up or down, depending on the air pressure signal (3-15 psi). The valve postioner can be of two types, (a) direct acting type and (b) feedback type. The direct acting type valve positioner is shown in Fig.6. Here the control pressure creates a downward pressure on the diaphragm against the spring, and the stem connected to the diaphragm moves up or down depending on the control pressure pc. At equilibrium the displacement of the stem can be expressed as: pc A = K x (7) where A is the area of the diaphragm and K is the spring constant. But the major shortcoming of this type of positioner is the nonlinear characteristics. Though ideally, the stem displacement is proportional to the control pressure (from (7)), the effective area of the diaphragm changes as it deflates. The spring characteristics is also not totally linear. Moreover, in (7) we have neglected the upward thrust force exerted by the fluid. The change in thrust force also causes the change in performance of the positioner. Besides the force exerted on the control valve is also not sufficient for handling valves for controlling large flow. As a result, the use of direct acting type valve positioner is limited to low pressure and small diameter pipelines.
Air

Diaphragm Spring

Stem

Flow

Plug Fig. 6 Direct acting type valve positioner

Version 2 EE IIT, Kharagpur 10

Pressure (3-15psi)

Spring Diaphragm

To valve

Feedback link

vent 20psi vent

Power cylinder Fig. 7 Feedback type valve positioner

The feedback type valve positioner (Fig.7) has a pilot cylinder with which the diaphragm is attached. The piston of this pilot cylinder opens or closes the air supply and vent ports to the main cylinder whose piston is connected to the stem of the control valve (not shown). There is a mechanical link connected to the stem that adjusts the fixed end of the spring connected to the diaphragm. This link provides the feedback to the postioner. As the control pressure increases, the diaphragm moves down, so is the piston of the pilot cylinder. This causes the lower chamber of the main cylinder to be connected to the 20 psi line and the upper chamber to the vent line. Compressed air enters the bottom of the main cylinder and the piston moves up. As the piston moves up, the feedback link compresses the spring further and this causes the diaphragm to move back to its original position. The air supply and the vent ports are now closed and the piston of the main cylinder remains at its previous position. The relationship between the control pressure and movement of the stem in this case is more or less linear. Moreover due to presence of power cylinder, the scheme is more suitable to position large control valves.

Version 2 EE IIT, Kharagpur 11

Conclusion
In this lesson we have discussed the construction and principle of operation of a number of pneumatic components normally used in a process control scheme. A flapper nozzle amplifier is most important component among these, the simplified characteristics of a flapper nozzle amplifier, assuming the air to be compressible has been presented in this lesson. The need for using air relay and feedback mechanism with a flapper nozzle amplifier is also elaborated. Majority of the valve positioners are pneumatic. Different types of pneumatic valve postioners are also discussed in this lesson. However two important pneumatic components have been left out. The first one is air pressure regulator and the second one is air filter. Air pressure regulator is needed to provide constant pressure air supply (20 psi) irrespective of air flow variation. Air filter removes moisture and dirt present in the air before it is used in the pneumatic components. Interested readers are requested to consult the books referred for understanding the construction and principle of operation of these two devices. Pneumatic controllers, though not so popular nowadays, are built up on these basic components discussed in this lesson. The details of pneumatic P-I-D controllers would be discussed in the next lesson.

References
1. 2. 3. 4. 5. 6. D.R. Coughanowr: Process systems analysis and control (2/e), McgrawHill, NY, 1991. D.P. Eckman: Automatic process Control, Wiley Eastern, New Delhi, 1958. B. Liptak: Process Control: Instrument Engineers Handbook W.L. Luyben and M.L. Luyben: Essentials of Process Control, McgrawHill, NY, 1997. P. Harriott: Process Control, Tata-McGrawHill, New Delhi, 1991. J.P. Bentley: Principles of Measurement Systems (3/e), Longman, U.K., 1995.

Review Questions
1. Explain with a simple sketch the principle of operation of a flapper nozzle amplifier. 2. Sketch the input-output characteristics of a flapper nozzle amplifier. 3. Identify the factors those affect the sensitivity of a flapper nozzle amplifier. 4. What is the function of air relay in pneumatic control? 5. What is the major advantage of using a flapper nozzle amplifier in closed loop? 6. Sketch and explain the working principle of a pneumatic torque balance transducer. 7. Explain the construction and working principle of a direct acting type pneumatic valve postioner. What are the limitations of this type of positioners? 8. How can you convert a 4-20mA current signal to a 3-15 psi pressure signal? Explain with a schematic.

Version 2 EE IIT, Kharagpur 12

Module 6
Actuators
Version 2 EE IIT, Kharagpur 1

Lesson 30
Pneumatic Control Systems
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
At the end of this lesson, the student should be able to • • • • • • Sketch the schematic diagram of a pneumatic proportional controller. Apply linearisation technique to develop the transfer function of a pneumatic proportional controller. Identify the major difference in construction among pneumatic P, P-D and P-I controllers. Identify the varying element by which the proportional gain of a P-controller can be adjusted. Identify the varying elements for adjusting the derivative and integral times in P-D and PI controllers. Develop the transfer function of a pneumatic P-D controller.

Introduction
In the last lesson we have discussed on the construction and principles of operation of various pneumatic components. In this lesson, we shall try to understand how these components can be combined to make a complete pneumatic control system. We shall particularly concentrate on the working of various PID controllers. These controllers together with the final control elements (diaphragm type valve positioners) provide the essentials of a pneumatic control system. Pneumatic controllers are the earliest type of controllers used in industry and still find regular use in many applications. But probably the more interesting part of pneumatic control is its principle of operation and how the derivative and integral parts can be generated by simply throttling the valves in the air line. The implementation issues of PID controllers have been discussed in Lesson 14. The details of the implementation of a pneumatic PID controller have been elaborated in this lesson.

Pneumatic Proportional Controller
Consider the pneumatic system shown in Fig.1. It consists of several pneumatic components discussed in Lesson 29. The components, which can be easily identified, are: flapper nozzle amplifier, air relay, bellows and springs, feedback arrangement etc. The overall arrangement is known as a pneumatic proportional controller. It acts as a controller in a pneumatic system generating output pressure proportional to the displacement e at one end of the link. The principle of operation is similar to that of Fig. 4 in lesson 29. The input to the system is a small linear displacement e and the output is pressure Po. The input displacement may be caused by a small differential pressure to apair of bellows, or by a small current driving an electromagnetic unit. There are two springs K2 and Kf those exert forces against the movements of the bellows A2 and Af. For a positive displacement of e (towards right) will cause decrease of pressure in the flapper nozzle. This will cause an upward movement of the bellows A2 (decrease in y). Consequently the output pressure of the air relay will increase. The increase in output pressure will move the free end of the feedback bellows towards left, bringing in the gap between the flapper and nozzle to almost its original value. We will first develop the closed loop representation of the scheme and from there the inputoutput relationship will be worked out. The air is assumed to be impressible here. Version 2 EE IIT, Kharagpur 3

e Input Displacement α Ps A2 K2 y P2 β x Pivot Af p0 z Feedback bellows Output P0 to system

Kf

Air Relay

Fig. 1 A pneumatic proportional controller The first problem we encounter in order to obtain the linearised relationship is the nonlinear characteristics of the flapper nozzle amplifier as shown in Fig.2. The problem can be circumvented by linearisng the characteristics over an operating point. Suppose Xi,o is the nominal gap between the flapper and the nozzle and the pressure at this operating point is P2,o. Any incremental change in displacement Xi (say x) will cause an incremental change in pressure of p2. The linearised relation ship can be obtained by taking into account of the first order term in Taylor series expansion as:

p2 =

∂P2,o

∂X i , o

x = − K1 x

(1)

The term

∂P2,o ∂X i , o

indicates the slope of the flapper nozzle characteristics at the operating point.

This slope can be taken as constant and is denoted here by − K1 , indicating a negative slope.

Version 2 EE IIT, Kharagpur 4

P2

Slope = -K1 Operating point P2, 0

0

Xi,0

Xi

Fig. 2 Characteristics of a flapper nozzle amplifier. Similarly, for the air relay we can obtain linearised characteristics as given by: po = − Ko y (2) indicating decrease in output pressure for positive (downward) y. Now the two bellows move against two springs. Let the constants of the springs be K2 and Kf. We assume the areas of the bellows be A2 and Af. Then, A (3) K 2 y = p2 A2 ; or, y = 2 p2 K2 Similarly, A K f z = po Af ; or, z = f po . (4) Kf In all the cases the variables represented by small letters denote the deviations from the values at the operating points. Now consider the movement of the link. It has two independent inputs (e and z). Applying superposition, the net displacement signal at the flapper nozzle is given by:

x=

β

α+β

e−

α

α+β

z

(5)

Now we can draw the block diagram of the closed loop feedback using the above expressions and as shown in Fig. 3 (a).This block diagram can be further simplified as shown in Fig. 3(b)-(d), where

Version 2 EE IIT, Kharagpur 5

e

β α+β

+ -

x

- K1

p2

A2 K2

- K0

po

α α+β
(a) e

z

Af Kf

β α

α α+β

+ -

K'

po

α α+β
(b)

Af Kf

e

β α

+ -

α α+β

K'

po

Af Kf
(c) e

β α

K'' (d)

po

Fig. 3 Block diagram of the closed loop system.

K ' = K1 K o
and

A2 K2

Version 2 EE IIT, Kharagpur 6

α
K '' =

α+β A α 1+ K' f Kf α+β If we assume Af α 〉〉 1 K' (6) Kf α+β then Kf , and K '' ≈ Af po β K f (7) ≈ e α Af It can be easily seen that the condition in (6) can be satisfied if the gains of the flapper nozzle amplifier and the air relay are high and the simplified relationship in (7) shows that po is proportional to e. Further the proportionality constant depends only on the area of the feedback bellows, the spring constant of the feedback bellows and the distances α and β only. The sensitivity is not dependent on the flapper nozzle and air relay characteristics as long as (6) is satisfied (their gains are indeed high, so it is not difficult to satisfy (6)). The proportional gain can be adjusted by adjusting the distances α and β .

K'

Pneumatic Proportional plus derivative controller
The schematic arrangement of the pneumatic proportional controller is shown in Fig.1. The proportional plus derivative (PD) action can also be generated in a pneumatic controller by introducing a restrictor in the line towards the feedback bellows in Fig. 1. This particular arrangement is shown separately in Fig. 4, all other parts remaining same as in Fig.1. The area of opening of the restrictor is small, so that the time constant associated with changing the pressure inside the feedback bellows is appreciable. In order to explain the generation of PD action, we need to study in detail the performance of the section shown in Fig. 4. Since air inside the feedback bellows is confined, the compressibility of air needs to be considered. e x Kf Af Pf z From air relay Fig. 4 Feedback arrangement for PD controller Version 2 EE IIT, Kharagpur 7 P0 to system Restrictor

Let Po and Pf are the pressures before and after the restrictor respectively. The mass flow rate of the fluid Gs through the restrictor is proportional to the square root of the pressure difference between Po and Pf . In general, we can write: Gs = f ( Po , Pf ) The above nonlinear expression can be linearised by considering the incremental changes g s , po and p f of the values of variables Gs , Po and Pf at the operating point as:

g s = C1 po + C2 p f
where, C1 = ∂Gs ,o and C2 = ∂Gs ,o

(8)

are the slopes of the curves Gs vs. Po and Gs vs. Pf at the ∂Po,o ∂Pf ,o operating point. These values can be obtained either experimentally, or from theoretical considerations. In fact, C1 = −C2 ; this can be ascertained from the fact that if Po and Pf both change from the operating point by the same amount (so that po − p f = 0 ) there is no change in the pressure drop, and so there will also be no change in mass flow rate and g s = 0 . From (8). We obtain C1 = −C2 and (8) can be rewritten as: g s = C1 ( po − p f ) (9) Further, the pressure Pf inside the feedback bellows can also be obtained from the expression: MRT f (10) Pf = Vf where M = Mass of the gas inside the bellows Vf = Volume of the gas inside the bellows Tf = Temperature of the gas (constant) Let m and vf be the changes in mass of the gas and volume of the gas from the operating point corresponding change in pressure by p f from the operating point. From (10), one can also obtain the linearised expression around the operating point as: ∂Pf ,o ∂Pf ,o pf = m+ vf ∂M o ∂V f ,o

= C3m − C4v f

(11)

where C3 and C4 are constants. The negative sign associated with C4 is due to the fact that increase in volume causes decrease in pressure. Now the change in volume inside the bellows is due to the displacement of the free end, and v f = Af z (12) Again the force balance condition at the feedback bellows gives: K f z = p f Af , or,
z Af Equating (11) and (13), one can obtain: pf = Kf

(13)

Version 2 EE IIT, Kharagpur 8

⎡Kf ⎤ + C4 Af ⎥ z ⎢ ⎢ Af ⎥ ⎣ ⎦ Differentiating the above equation, we obtain: ⎤ dz dm 1 ⎡ K f gs = = + C4 A f ⎥ . ⎢ dt C3 ⎢ Af ⎥ dt ⎣ ⎦ Again equating (9) and(15), ⎤ dz Kf 1 ⎡Kf + C4 Af ⎥ = po − p f = po − z (from (13)). ⎢ C1C3 ⎢ Af Af ⎥ dt ⎣ ⎦ m= 1 C3
Or,

(14)

(15)

K f ⎡ 1 ⎛ C4 A2 f ⎢ ⎜1 + Af ⎢ C1C3 ⎜ Kf ⎝ ⎣

⎤ ⎞ dz ⎟ + z ⎥ = po . ⎟ dt ⎥ ⎠ ⎦

Taking the Laplace Transformation, we can have: ( Af / K f ) p (s) , z (s) = o 1 + sτ d where, 1 + C4 A2 / K f f τd = C1C3

(16)

Compare (16) with the relationship obtained z and po in (4) in a proportional controller as: Af z= po . Kf It is clear that the introduction of the restrictor in the feedback bellows introduces a time constant in the feedback path. Further, by varying the restrictor area, C1 can be changed (refer (8)), thus changing the time constant τ d . Other parts in the block diagram for P-controller shown in Fig. 3 remains the same. For the sake of simplicity, let us assume the link lengths α = β . In that case we can develop the simplified block diagram for the system shown in Fig. 1 with the modified feedback bellows configuration shown in Fig. 4. The simplified block diagram can be expressed as shown in Fig.5, wherefrom we obtain,
+ z

e

K′

2

po

(1 + sτ d )
Fig. 5 Block diagram of the P-D controller. Version 2 EE IIT, Kharagpur 9

Af K f

e x Af, Kf PI z From Air relay Restrictor Po to system Fig. 6 Feedback arrangement for PI controller Af Po

K′ (1 + τ d s ) po K /2 2 = ( s) = e ⎛ K ′Af ⎞ ⎛ K ' Af ⎞ 1+ ⎜ ⎟ +τ d s ⎟ (1 + τ d s ) ⎜1 + ⎜ ⎜ 2K f ⎟ 2K f ⎟ ⎝ ⎠ ⎝ ⎠ Further, if we assume that, ( K ' Af ) /(2 K f )〉〉τ d and also 〉〉 1, then (17) simplifies to:
'

(17)

(18)

Kf po (s) ≈ (1 + τ d s ) e Af

(19)

that is similar to the transfer function of a standard P-D controller as: K p (1 + τ d s) . Thus by simple introduction of a restrictor in the line connecting the feedback bellows can transform the pneumatic P-controller in Fig. 1 to a P-D controller. Note that with a lagged feedback signal (as seen from the transfer function of the feedback block) the closed loop transfer function provides a net lead, with the same time constant as the lag time constant of the feedback path.

Pneumatic Proportional Plus Integral Controller
Proportional plus Integral (P-I) action can also be generated in a pneumatic controller in a similar way as discussed above for P-D controller, but by adding an additional feedback bellows with restrictor to the left of the link as shown in Fig. 6. There is no restriction in the air line connecting the bellows on the right hand side. There is also a spring surrounding the rhs bellows. We can also develop the equations here for the integrating bellows corresponding to (11)-(13) as: pI = C3m − C4vI (20)

vI = − Af z ( po − pI ) Af = k f z

(21) (22)

Version 2 EE IIT, Kharagpur 10

Proceeding in a similar way, one can obtain the expression relating z and po as: Af s C1C3 K f KI s z ( s) = po ( s ) po ( s ) 2 1+τ I s ⎛ 1 + C4 A f / K f ⎞ 1+ ⎜ ⎟s ⎜ ⎟ C1C3 ⎝ ⎠

(23)

The closed loop block diagram for P-D controller (shown in Fig.5) now gets modified for P-I controller as shown in Fig.7. The overall transfer function can be obtained as:
po (s) = e ( K ' / 2) ( K ' / 2)(1 + τ I s ) = ( K ' / 2) K I s K 'KI )s 1+ 1 + (τ I + 1+τ I s 2

(24)

If

K 'KI K 'KI s〉〉1 , then 〉〉τ I and 2 2 po 1 τ 1 (s) = (1 + τ I s) = I (1 + ) τIs e KI s KI

(25)

The above transfer function is clearly of a P-I controller. The time constant can be varied by varying the restrictor opening. The reader can now easily understand the action of a complete pneumatic P-I controller shown schematically in Fig. 4 of Lesson14. Here again, the feedback path dynamics (given in (23)) is approximately of a differentiator. The effect is inverted in the closed loop, generating integral action, having a reset time equal to the lag time constant of the feedback path.
+ z

e

K′

2

po

KI s (1 + τ I s )
Fig. 7 Block diagram of the P-I controller.

Conclusion
In this lesson we have discussed the working principles of pneumatic P, PD and PI controllers. PID action can also be generated by combining Fig. 4 and Fig.6. The major issues of a pneumatic controller we have learnt are (i) the proportional gain of a pneumatic proportional controller can be adjusted by adjusting the segment lengths α and β of the link, (ii) the proportional gain is independent of the flapper nozzle and relay gains, (iii) the derivative and integral actions can be generated by putting restrictors in the feedback bellows and their time constants τ d and τ I can be adjusted by throttling the restrictors. While deriving the working relationship of these controllers, we have also learnt how to linearise the characteristics of a nonlinear linear device. This Version 2 EE IIT, Kharagpur 11

knowledge reminds us that all the variables in the transfer functions developed are actually the changes from the nominal operating points and the linearised transfer function is valid for small deviations from operating points only. Pneumatic control system is free from electromagnetic hazards and the problem of short circuit; thus can be used as an intrinsically safe system where the ambient may contain highly combustible vapours as in a petrochemical and natural gas processing unit. The major limitation of this type of systems are that they are slow in response. Regular maintenance for keeping the control elements free from dust and moisture is also necessary fro proper operation of the controllers.

References
1. 2. 3. F.H. Raven: Automatic Control Engineering (4/e), McGraw-Hill, NY, 1987. D.P. Eckman: Automatic process Control, Wiley Eastern, New Delhi, 1958. K. Ogata: Modern Control engineering (2/e), Prentice Hall of India, New Delhi, 1995.

Review Questions
1. 2. 3. 4. 5. What is the function of air relay in Fig.1? Sketch the schematic arrangement of a pneumatic proportional controller and draw the closed loop block diagram. A high gain feedback system reduces the effect of parameter variations on the performance. Can you justify the statement with a pneumatic proportional controller? Identify the element that you should vary for adjusting the derivative time τ d . A nonlinear expression: Y = 2 X 12 X 2 is linearised over the operating point (X1 =1, X2 =1) and linearised expression is given by y = C1 x1 + C2 x2 (where y, x1 and x2 are the deviations from the operating point). Find the values of C1 and C2.

Answer
Q5. C1= 4, C2=2.

Version 2 EE IIT, Kharagpur 12

Module 7
Electrical Machine Drives
Version 2 EE IIT, Kharagpur 1

Lesson 31
Energy Savings with Variable Speed Drives
Version 2 EE IIT, Kharagpur 2

Lesson Objectives
• • • To describe typical methods of flow control by industrial fans and pumps To be able to determine operating points from pump/fan and load characteristics To demonstrate energy saving with variable speed drive method of flow control compared to throttling

Introduction
The AC induction motor is the major converter of electrical energy into mechanical and other useable forms. For this purpose, about two thirds of the electrical energy produced is fed to motors. Much of the power that is consumed by AC motors goes into the operation of fans, blowers and pumps. It has been estimated that approximately 50% of the motors in use are for these types of loads. These particular loads — fans, blowers and pumps, are particularly attractive to look at for energy savings. Several alternate methods of control for fans and pumps have been advanced recently that show substantial energy savings over traditional methods. Basically, fans and pumps are designed to be capable of meeting the maximum demand of the system in which they are installed. However, quite often the actual demand could vary and be much less than the designed capacity. These conditions are accommodated by adding outlet dampers to fans or throttling valves to pumps. These control methods are effective, inexpensive and simple, but severely affect the efficiency of the system. Others forms of control are now available to adapt fans and pumps to varying demands, which do not decrease the efficiency of the system as much. Newer methods include direct variable speed control of the fan or pump motor. This method produces a more efficient means of flow control than the existing methods. In addition, adjustable frequency drives offer a distinct advantage over other forms of variable speed control.

Fans: Characteristics and Operation
Large fans and blowers are routinely used in central air conditioning systems, boilers, drives and the chemical operations. The most common fan is the centrifugal fan that imparts energy in to air by centrifugal force. This results in an increase in pressure and produces air flow at the outlet of the fan.

Version 2 EE IIT, Kharagpur 3

Fig. 31.1 is a plot of outlet pressure versus the flow of air of a typical centrifugal fan at a given speed. Standard fan curves usually show a number of curves for different fan speeds and include the loci of constant fan efficiencies and power requirements on the operating characteristics. These are all useful for selecting the optimum fan for any application. They also are needed to predict fan operation and other parameters when the fan operation is changed. Appendix 1 gives an example of a typical fan curve for an industrial fan. Fig. 31.2 shows a typical system pressure-flow characteristics curve intersecting a typical fan curve.

Version 2 EE IIT, Kharagpur 4

The system curve shows the requirements of the vent system that the fan is used on. It shows how much pressure is required from the fan to overcome system losses and produce airflow. The fan curve is a plot of fan capability independent of a system. The system curve is a plot of "load" requirement independent of the fan. The intersection of these two curves is the natural operating point. It is the actual pressure and flow that will occur at the fan outlet when this system is operated. Without external control, the fan will operate at this point. Many systems however require operation at a wide variety of points. Fig. 31.3 shows a profile of the typical variations in flow experienced in a typical system.

There are several methods used to modulate or vary the flow to achieve the optimum points. Apart from the method of cycling, the other methods affect either the system curve or the fan curve to produce a different natural operating point. In so doing, they also may change the fan's efficiency and the power requirements. Below these methods are explained in brief.

Points to Ponder: 1
A. Why is it that typically load requirements are stated in terms of flow rates? B. Why is it that the load requirements are generally show a parabolic pressure flow characteristics? Version 2 EE IIT, Kharagpur 5

On-Off Control
This is typically done in home heating systems and air conditioners. Here, depending on temperature of the space in question and the desired temperature setting, the fan is switched on and off, cyclically. Although the average temperature can be maintained by this method, this produces erratic airflow, causes temperature to oscillate and is generally unacceptable for commercial or industrial use.

Points to Ponder: 2
A. Why is this system acceptable for home HVAC (the acronym for Heating Ventilation and Air Conditioning)? B. Is it energy efficient?

Outlet Dampers
The outlet dampers affect the system curve by increasing the resistance to air flow. The system curve is a simple function that can be stated as P = K × Q2, where P is the pressure required to produce a given flow Q in the system. K is a characteristics of the system that represents the resistance to airflow. For different values of outlet vane opening, different values of K are obtained. Fig. 31.5 shows several different system curves indicating different outlet damper positions.

Version 2 EE IIT, Kharagpur 6

The power requirement can be derived from computing the rectangular areas shown in Fig. 31.5 at any operating point. Figure 31.7 shows the corresponding variations in power requirement for this type of operation. From the figure it can be seen that the power decreases gradually as the flow is decreased.

Version 2 EE IIT, Kharagpur 7

Points to Ponder: 3
A. Can you identify, which of three system curves shown in Fig. 31.5 correspond to the maximum opening of the output damper? B. Assume that the pressure-flow curve C shown in Fig. 31.5 represent the characteristics of the output element f the system. Then, a. Find the system efficiency at 60% of rated flow. b. Does the efficiency increase, or decrease with flow rate?

Variable Speed Drive
This method changes fan curve by changing the speed of the fan. For a given load on the fan the pressures and flows at two different speeds, N1 and N2 are given as follows:

Q2 N 2 = Q1 N1

P2 ⎛ N 2 ⎞ =⎜ ⎟ P1 ⎝ N1 ⎠

2

W2 ⎛ N 2 ⎞ =⎜ ⎟ W1 ⎝ N1 ⎠

3

where N= Fan Speed, Q= Volume Flow Rate, P= Pressure, W= Power. Note that eliminating speed from the above equations gives back the law, P = K×Q2 for the system. Thus, by this method, the operating point for a given system load shifts along the system characteristic curve as the speed of the fan is varied. Fig. 31.7 is a representation of the variable speed method.

Fig. 31.7 shows the significant reduction in horsepower achieved by this method. Thus, in this method a desired amount of flow is achieved with the minimum of input power. The other two Version 2 EE IIT, Kharagpur 8

methods modify some system parameter, which generally results in the reduction of efficiency of the fan. This is why the power demand is greater than the variable speed method.

Points to Ponder: 4
A. Can you identify, which of three system curves shown in Figure 31.8 correspond to the minimum speed of the pump? B. Assume that the pressure-flow curve C shown in Figure 31.6 represent the characteristics of the output element f the system. Then, a. Find the system efficiency at 60% of rated flow. b. Does the efficiency increase, or decrease with pump speed?

Energy Savings by Different Flow Control Methods Outlet Damper
In this section the power consumption for two of the above methods, namely the variable speed method and the outlet damper method, and their associated costs of operation are estimated for a given load profile and a fan curve (shown in Appendix 1). Assume the fan selected has a rated speed of 300 RPM and 100% flow is to equal 100,000 CFM as shown on the chart. Assume the following load profile. Flow (cfm) Duty Cycle (% of time)

100% 80% 60% 40%

10% 40% 40% 10%

For each operating point, one can obtain the required power from the fan curve by locating the corresponding fan pressure. Note that, at all these operating points the speed is assumed constant. This power is multiplied by the fraction of the total time, for which the fan operates at this point. These "weighted horsepowers" are then summed to produce an average horsepower that represents the average energy consumption of the fan. Flow(cfm) 100 80 60 40 Duty Cycle 10 40 40 10 Power (hp) 35 35 31 27 Weighted Power(hp) 3.5 14.0 12.4 2.7 Total 32.6

Version 2 EE IIT, Kharagpur 9

Variable Speed Drive
To assess the energy savings, similar calculations are to be carried out to obtain an average horsepower for variable speed operation. The fan curve does not directly show the operating characteristics at varying speeds. However, these can be obtained using the laws of variation of pressure and flow with speed. From the fan curve, 100% flow, (say Q1) at 100% speed, (say N1) requires 35 HP. Now since, Q2/Q1 = N2/N1, the new value of speed N2 required to establish Q2 can be obtained easily. This value of N2, substituted into the power formula, (P2/ P1) = (N2/N1)3 would then yield the new value of P2 needed to establish Q2 at speed N2. When Q1 = 100% and W1 = 35 HP, the values of W2 for various values of Q2 are shown below. Q2
W2

80
18

60
7.56

40
2.24

These calculated values do match the points available on the fan curve. Now it is possible to calculate the average horsepower. Flow(cfm) 100 80 60 40 Duty Cycle 10 40 40 10 Power (hp) 35 18 7.56 2.24 Weighted Power(hp)

3.5 7.2 3.024 0.224 TOTAL 13.948 Comparing the above figure with that calculated for the outlet damper method indicates the difference in energy consumption. The variable speed method requires less than half the energy of the outlet damper method (based on the typical duty cycle). As an example of the cost difference between these methods, let's assume the system operates twenty-four hours per day (730 hours per month), and the cost of electricity is Rs. 2.00 per kilowatt-hour. The cost of electricity is determined in terms of the energy in kilowatt hour per month. WEIGHTED HORSEPOWER KW/HP HR/MONTH KWH/MONTH COST (Rs./KWh) TOTAL COST(Rs.) OUTLET DAMPER 32.6 .746 730 17,753 2.00 35,506 VARIABLE SPEED 13.948 .746 730 7,596 2.00 15,192

There is over a 20,000 Rs. per month savings available by using the variable speed method instead of the outlet damper method.

Version 2 EE IIT, Kharagpur 10

This example only takes into account the operation of the fan. In practice the motor efficiency and the drive efficiency should also be taken into account. It is impractical to list or chart the motor and drive efficiencies in this lesson for all possible load conditions that could occur. However, since the same motor would be used in both examples shown here, the difference in motor efficiencies would be minimized and not significantly affect the results shown.

Points to Ponder: 5
A. Is energy saving the only factor to determine which technology to adopt in a given industrial situation? If not, mention at least three other issues that also need to be considered, let us say, to decide whether to go for a throttling or a variable speed drive flow control system ? B. How does the flow-demand profile affect the extent of energy savings? For example, would the energy saving be larger if the ratio of the minimum demand to maximum demand be high? How do the durations of these demand affects the saving?

Pumps: Characteristics and Operation
Pumps are generally grouped into two broad categories, positive displacement pumps and centrifugal pumps. Positive displacement pumps use mechanical means to vary the size (or move), the fluid chamber to cause the fluid to flow. Centrifugal pumps impart a momentum in the fluid by rotating impellers immersed in the fluid. The momentum produces an increase in pressure or flow at the pump outlet. The vast majority of pumps used today are of the centrifugal type. Only centrifugal pumps are discussed here. Fig. 31.8 again shows two independent curves. One is the pump curve which is solely a function of the physical characteristics of the pump. The other curve is the system curve. This curve is completely dependent on the size of pipe, the length of pipe, the number and location of elbows, etc. Where these two curves intersect is called the natural operating point. That is where the pump pressure matches the system losses and everything is balanced. Note that this balance only occurs at one point (or at least should for stable system operation). If that point does occur at or at least come close to the desired point of operation, then the system is acceptable. If it does not come close enough then either the pump or the system physical arrangement has to be altered to correct to the desired point.

The following laws govern, similar to the case of fans, the operation of centrifugal pump characteristics at various pump speeds. 2 3 Q 2 N 2 P2 ⎛ N 2 ⎞ W2 ⎛ N 2 ⎞ = =⎜ =⎜ ⎟ ⎟ Q1 N1 P1 ⎝ N1 ⎠ W2 ⎝ N1 ⎠ where, N= Pump Speed Q= Flow P= Pressure W= Horsepower

Version 2 EE IIT, Kharagpur 11

Flow Control
Similar to the case of fans, there are two main methods of flow control in pumps, namely, use of a control or throttling valve and variable speed control of the pump. In view of the similarity, these are described in brief below.

Throttling
Consider a throttling system shown in Fig. 31.9. Two conditions of the system curve are shown, one with the valve open and the other with the valve throttled or partially closed. The result is that when the flow in the system is decreased, the pump head increases.

Variable Speed Drive
In comparison, the variable speed method takes advantage of the change in pump characteristics that occur when the pump impeller speed is changed. The new pump characteristics can be predicted from the laws stated earlier. In this method is that the pump head decreases as the flow Version 2 EE IIT, Kharagpur 12

is decreased. Fig. 31.10 on the following page gives an example of the system controlled from a variable speed pump.

Version 2 EE IIT, Kharagpur 13

Static Head
Static head is the pressure required to overcome an elevation change in the system. To get water from the base to a spout at the top of a vessel would cause a static head on this pump. A system with a static head does change the system curve and the horsepower requirements will change from that shown previously. Fig. 31.11 shows the system curves for systems with different static heads. Fig. 31.12 shows the horsepower requirements for each system. Version 2 EE IIT, Kharagpur 14

The system corresponding to the curve A is without static head. The system corresponding to the curve B requires a static head. The system corresponding to the curve C requires a double the static head and still has the same operations with the pump. Fig. 31.12 shows the power curves for the variable speed operations for the three systems A, B, and C. Curve D corresponds to throttle control. The dynamic operation of the throttling system does not change with static head. The static head does part of the work for the throttling valve. However, note that the horsepower requirement for this method remains above that of the variable speed method.

Version 2 EE IIT, Kharagpur 15

Lesson Summary
In this lesson we have presented the main methods for flow control with special focus on energy savings. Specifically, the following have been discussed. A. Major methods of flow control, such as, by output dampers and variable speed drives. B. Assessment of energy requirements by the two methods in a practical example and comparative energy savings by the variable speed method. C. Major methods of flow control by pumps, such as, by flow control valves and variable speed drives. D. Assessment of energy requirements by the two methods in a practical example and comparative energy savings by the variable speed method. E. Analysis of the effect of static heads in flow control of pumps. Version 2 EE IIT, Kharagpur 16

Answers, Remarks and Hints to Points to Ponder Points to Ponder: 1
A. Why is it that typically load requirements are stated in terms of flow rates? Ans: Major applications for fans, blowers and pumps are for cooling, combustion, heat exchange, as feed mechanisms for chemical process etc. Notice that in all these applications the volumetric flow rate of the fluid is what needs to be controlled. B. Why is it that the load requirements generally exhibit a parabolic pressure flow characteristics? Ans: Because that is the characteristics of pressure drops whenever turbulent flow occurs, by Bernoulli’s principle. Δ P ∝ V 2 ∝ Q2

Points to Ponder: 2
A. Why is this system acceptable for home HVAC? Ans: Because the large thermal capacitance of home spaces is high compared to the heat loads. Therefore, even with on-off control, temperature oscillations are small. Finally, the cost saving is acceptable to customers in comparison with comfort requirements which are non critical. B. Is it energy efficient? Ans: It is more energy efficient compared to the outlet damper method. However it has some disadvantages compared to the variable speed drive method. Firstly the continuous start-stop process causes additional energy loss. It also causes maintenance problems. For these reasons, on-off control is not employed in large pumps and fans.

Points to Ponder: 3
A. Can you identify, which of three system curves shown in Fig. 31.6 correspond to the maximum opening of the output damper? Ans: Curve A, because it has the lowest pressure drop for a given flow. B. Assume that the pressure-flow curve C shown in Fig. 31.6 represent the characteristics of the output element f the system. Then, a. Find the system efficiency at 60% of rated flow. Ans: Assume that in the fully open position, the pressure drop across the valve is zero, so that, the characteristics of the output element is given by Curve A. From curve C, the pressure drop across the output element is about 45%. Therefore the drop across Version 2 EE IIT, Kharagpur 17

the output element is about 85%. Thus the efficiency of the system at 60% flow is about 65%. b. Does the efficiency increase, or decrease with flow rate? Ans: The efficiency of the system obviously increases with flowrate. Check for the efficiencies for 80% and 100% flows.

Points to Ponder: 4
A. Can you identify, which of three system curves shown in Fig. 31.8 correspond to the maximum speed of the pump? Ans: Curve A, because it has the highest flow for a given pressure and the highest pressure for a given flow. That is consistent with the laws. B. Assume that the pressure-flow curve C shown in Fig. 31.8 represent the characteristics of the output element f the system. Then, a. Find the system efficiency at 60% of rated flow. Ans: The efficiency is always 100 %, if we ignore the pipe losses and the efficiency of the motor and the drive. This is because there is never any pressure drop across a damper, and so the whole of the pump power is deployed in the load. b. Does the efficiency increase, or decrease with flow rate? Ans: The efficiency of the system remains more or less constant, notwithstanding the efficiency of the motor and the drive which are likely to fall to some extent with speed.

Points to Ponder: 5
A. Is energy saving the only factor to determine which technology to adopt in a given industrial situation? If not, mention at least two other issues that also need to be considered, let us say, to decide whether to go for a throttling or a variable speed drive flow control system ? Ans: Two other issues are: i. Equipment cost ii. Maintenance cost B. How does the flow-demand profile affect the extent of energy savings? For example, would the energy saving be larger if the ratio of the minimum demand to maximum demand be high? How do the durations of these demand affects the saving? Ans: The energy saving of the variable speed drive system is realized if the ratio of the maximum to minimum loads is high and also if the system operates at less than maximum load for a significant fraction of time. Version 2 EE IIT, Kharagpur 18

Version 2 EE IIT, Kharagpur 19

Module 7
Electrical Machine Drives
Version 2 EE IIT, Kharagpur 1

Lesson 32
Step Motors: Principles, Construction and Drives
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
At the end of this lesson, the student should be able to • • • • • • • • Explain how a step motor is different from a conventional motor. Identify the major constructional difference between a permanent magnet and variable reluctance type motor. Distinguish between the terms full stepping and half stepping. Develop the switching sequence for a given step motor according to given requirements. Calculate the step angle. Explain what is meant by static position error. Name two different modes of operation for continuous rotation. Explain with schematic diagrams, open loop and closed loop control schemes used for step motors.

Introduction
Step motors (often referred as stepper motors) are different from all other types of electrical drives in the sense that they operate on discrete control pulses received and rotate in discrete steps. On the other hand ordinary electrical a.c and d.c drives are analog in nature and rotate continuously depending on magnitude and polarity of the control signal received. The discrete nature of operation of a step motor makes it suitable for directly interfacing with a computer and direct computer control. These motors are widely employed in industrial control, specifically for CNC machines, where open loop control in discrete steps are acceptable. These motors can also be adapted for continuous rotation. In this lesson we would discuss about the construction and principle of operations of different type of step motors and elaborate on the drive schemes used.

Construction
Step motors are normally of two types: (a) permanent magnet and (b) variable reluctance type. In a step motor the excitation voltage to the coils is d.c. and the number of phases indicates the number of windings. In both the two cases the excitation windings are in the stator. In a permanent magnet type step motor the rotor is a permanent magnet with a number of poles. On the other hand the rotor of a variable reluctance type motor is of a cylindrical structure with a number of projected teeth.

Permanent magnet step motor
The principle of step motor can be understood from the basic schematic arrangement of a small permanent magnet step motor is shown in Fig.1. This type of motor is called a two-phase twopole permanent magnet step motor; the number of windings being two (phase 1 and phase 2) each split into two identical halves; the rotor is a permanent magnet with two poles. So winding A is split into two halves A1 and A2. They are excited by constant d.c. voltage V and the direction of current through A1 and A2 can be set by switching of four switches Q1, Q2, Q3 and Version 2 EE IIT, Kharagpur 3

Q4 as shown in Fig.2(a). For example, if Q1 and Q2 are closed, the current flows from A1 to A2, while closing of the switches Q3 and Q4 sets the direction of current from A2 to A1. Similar is the case for the halves B1 and B2 where four switches Q5-Q8 are used to control the direction of current as shown in Fig. 2(b). The directions of the currents and the corresponding polarities of the induced magnets are shown in Fig. 1. Phase 1

Phase 2 N N B1 S N

A1

B2 A2

Fig. 1 Schematic diagram of a two-phase two-pole permanent magnet stepper motor. Switching Sequence + Q1 Q3 A1 A2 B1 Fig. 2 Switching sequence for Fig. 1. Now consider Fig. 3. Let Winding A be energised and the induced magnetic poles are as shown in Fig. 3(a) (we will denote the switching condition as S1=1). The other winding B is not energised. As a result the moving permanent magnet will align itself along the axis of the stator poles as shown in Fig. 3(a). In the next step, both the windings A and B are excited simultaneously, and the polarities of the stator poles are as shown in Fig. 3(b). We shall denote S2=1, for this switching arrangement for winding B. The rotor magnet will now rotate by an angle of 45o and align itself with the resultant magnetic field produced. In the next step, if we now make S1=0 (thereby de-energising winding A), the rotor will rotate further clockwise by 45o and align itself along winding B, as shown in Fig. 3(c). In this way if we keep on changing the switching sequence, the rotor will keep on rotating by 45o in each step in the clockwise direction. The switching sequences for the switches Q1 to Q8 for first four steps are tabulated in Table 1. B2 V Q2 Q4 Q5 Q7 Q6 + V Q8

Version 2 EE IIT, Kharagpur 4

Step 1

S1 = 1

Step 2

S1 = 1

N Pole S S2 = 0 N S Pole S2 = 1 S Pole

N Pole S N S Pole N Pole

(a)

(b)

Step 3

S1 = 0

Step 4

S1 = -1

S Pole S Pole S2 = 1 N S N Pole S2 = 1 S Pole N S N Pole N Pole

(c)

(d)

Fig. 3 Stepping sequence (half-stepping) for a two-phase two-pole PM step motor for clockwise rotation. Table 1 Switching sequence corresponding to the movement shown in Fig.3 Step 1 Step 2 Step 3 Step 4 Q1-Q2 Q3-Q4 Q5-Q6 Q7-Q8 ON (S2=1) ON (S2=1) ON (S2=1) ON (S1=1) ON (S1=1) ON (S1= -1)

Version 2 EE IIT, Kharagpur 5

It is apparent from Table 1 and Fig. 3 that for this type of switching the step angle is 45o and it takes 8 steps to complete a complete revolution. So we have 8 steps / revolution. It can also be seen from Table 1 that a pair of switch (say Q7-Q8) remains closed during consecutive three steps of rotation and there is an overlap at every alternate step where both the two windings are energised. This arrangement for controlling the step motor movement is known as half stepping. The direction of rotation can be reversed by changing the order of the switching sequence. It is also possible to have an excitation arrangement where each phase is excited one at a time and there is no overlapping where both the phases are energised simultaneously, though it is not possible for the configuration shown in Fig.1, since that will require the rotor to rotate by 90o in each step and in the process, may inadventedly get locked in the previous position. But full stepping is achievable for other cases, as for example for the two-phase six-pole permanent magnet step motor as shown in Fig. 4. In this case, the stator pitch θ s = 900 and the rotor pitch

θ r = 600 ; the full step angle is given by

θ f s = θ s ∼ θ r = 300

and the half step angle

θ h s = (θ s ∼ θ r ) / 2 = 150 .
the sequence of switching.

The desired direction of rotation can be achieved by choosing

A1
S N

N

S

B1

S

N

B2 A2

Fig. 4 Two-phase six-pole permanent magnet stop motor. The advantage of a permanent magnet step motor is that it has a holding torque. This means that due to the presence of permanent magnet the rotor will lock itself along the stator pole even when the excitation coils are de-energised. But the major disadvantage is that the direction of current for each winding needs to be reversed. This requires more number of transistor switches that may make the driving circuit unwieldy. This disadvantage can be overcome with a variable reluctance type step motor, as explained in the next section. Another way of reducing the number of switches is to use unipolar winding. In unipolar winding, there are two windings per pole, out of which only one is excited at a time. The windings in a pole are wound in opposite direction, thus either N-pole or S-pole, depending on which one is excited.

Version 2 EE IIT, Kharagpur 6

Variable Reluctance type Step Motor
Variable reluctance type step motors do not require reversing of current through the coils, but at the same time do not have any holding torque. Compared to permanent magnet step motors, their step angles are also much smaller. Step angle as low as 1.8o can be achieved with this type of motors. Here the rotor is a cylindrical soft iron core with projected teeth. When a particular stator coil is excited, the rotor aligns itself such that one pair of teeth is along the energised stator coil, at the minimum reluctance path. The schematic arrangement of a three phase VR motor with 12 stator poles (teeth) and eight rotor teeth is shown in Fig.5. When phase-1 is energised, the rotor will align itself as shown in the figure. In the next step, if phase-1 is switched off and phase-2 is switched on, the rotor will rotate in CCW direction by an angle of 15o. This can be understood from the following derivation: Here the stator pole pitch 3600 3600 θs = = = 300 . Number of stator pole teeth 12 On the other hand, the rotor pole pitch 3600 3600 θr = = = 450 Number of rotor pole teeth 8 So the full step angle θ f s

= θ s ∼ θ r = 450 - 300 = 150 .

Similar to the earlier case, we can also have half stepping where step angle of 7.5o can be achieved. Te switching sequence for rotation in the counter clockwise direction with half stepping would be 1-(1,2)-2-(2,3)-3-(3,1)-1…. Further reduction of step angle is possible by increasing the number of stator and rotor teeth. Besides, multi-stack stators are also used for achieving smaller step angle, where there are several stacks of stator windings skewed from each other by a certain angle. It has been already mentioned that the VR motors do not have any holding torque. It is natural because, when the stator coils are de-energised there is no magnetic force present and the rotor is free. Hybrid step motors are improved versions of single stack VR motors, where the basic constructions are modified slightly in order to achieve holding torque. However this part will not be discussed in this lesson. Interested readers may consult the books given in the reference
3 2 1 S 3 2 N 1 3 1 N 2 3 S 1 2

Fig. 5 Three-phase single-stack VR step motor with twelve stator poles (teeth) and eight rotor teeth. Version 2 EE IIT, Kharagpur 7

Typical specification of a step motor
4 phase permanent magnet unipolar stepper motor. Supply voltage:12V, Power: 5W, Current: 400 mA. 200 steps/revolution, (full step angle-1.8o, half step angle- 0.9o). Resistance/ phase: 30 Ω , Inductance/ phase: 23 mH. Holding torque: 2000 gm-cm, Rotor inertia: 22 gm-cm2. Shaft diameter: 5cm.

Driving Circuit
The switches shown in Fig.2 are essential for energising the coils of the step motor. In present day applications, semiconductor switches are universally employed fro this purpose. A typical transistor switch arrangement is shown in Fig.6(a). A positive going pulse will turn on the transistor and energise the coil. When the base voltage is withdrawn, the current through the coil stops. A free wheeling diode is normally provided across the coil so that when the current through the coil is suddenly withdrawn, the large induced emf finds a return path for current and its magnitude is reduced. Note that the developed torque would always be restorative, i.e. of signal so that it tends to bring back the rotor to equilibrium, i.e. θ = 0 . Now suppose, we give a rectangular pulse to the base of the transistor. Ideally the current through the coil will be as shown by the dotted lines in Fig. 6(b). But the current cannot rise instantaneously due to the presence of inductor and the actual current pulse will be as shown by the continuous line in Fig. 6(b). As a result sufficient torque will be generated only when the current reaches the steady state value. Thus the time constant (T0) of the winding limits the frequency of switching. For proper operation, the width of each pulse should be at least 6-8 times the time constant of the winding. +

Coil 95% Pulse

3T0 (a) (b)

Fig. 6 (a) Driving circuit of a coil (b) Current pulse: line – ideal, dotted - actual

Static Torque Curve

Version 2 EE IIT, Kharagpur 8

The torque developed in a step motor is not constant and is position dependent. To understand the torque vs. displacement curve, consider a simple three-phase VR type motor with a rectangular rotor as shown in Fig. 7(a). Suppose Phase-1 is excited and the rotor is aligned with its axis ( θ = 0 ). At this position the rotor is not experiencing any torque (otherwise it should have moved, friction is neglected). With phase-1 remaining excited if we now rotate the rotor by hand through ±900 , it will experience torque as shown in Fig. 7(b). The torque is zero when θ = ±900 . In between the torque direction will be +ve or –ve depending on the direction of θ , but the magnitude is not constant. However the zone of operation for phase-1 is limited to ±600 , since phase-2 or phase-3 takes up on either side. For clockwise operation, suppose the initial rotor position was –60o. If phase-1 is excited, it will bring the rotor to the position θ = 0 . If phase-2 is now excited the rotor will rotate clockwise by +60o and settle there. Phase-1

Torque Operating zone Phase-2

Phase-3 Phase-3 Phase-2 - 90º - 60º Phase-1 θ = 0º (a) (b) + 60º

0

+ 90º θ

Fig. 7 (a) Three phase two pole VR-type step motor (b) Its torque vs displacement characteristics corresponding to phase-1. Subsequently phase-3 takes up and the rotor rotates further to +1200. The overall torque vs. displacement characteristics is shown in Fig. 8. Now the number of rotor poles considered above is two. It should be noted that if the number of rotor poles were four, the range of angle for which torque curve will cover a full cycle will reduce to ±450 and the operating zone for each phase will reduce to Δθ = 300 for clockwise rotation, where Δθ is the full step angle. Thus the operation of each phase in the torquedisplacement characteristics in Fig. 8 is shown to be Δθ . But the curve corresponding to excitation of only one phase for each phase will be sinusoidal in nature.

Version 2 EE IIT, Kharagpur 9

Static Torque T Phase 1 Phase 2 Phase 3 Phase 1

TL 0 -θe ∆θ 2.∆θ 3.∆θ Angle θ

Fig. 8 Periodicity of the single-phase static torque distribution (a three-phase example).

Static Position Error
Consider the static torque characteristics corresponding to phase –1, as shown in Fig. 9. Under ideal condition, the equilibrium position in Fig. 9 is zero. However if there is a static load torque TL is present, the equilibrium point will shift to −θ e , since beyond this point the torque generated will not be sufficient to overcome the static load torque. Thus there will be a static position error of θ e . Assuming the torque angular displacement curve is sinusoidal, it can be shown that:
p −1 ⎛ TL ⎞ (1) sin ⎜ ⎟ n ⎝ Tmax ⎠ where p = number of phases, n = number of steps/ revolution and Tmax is the maximum torque generated.. Thus more the number of steps/ revolution (smaller step angle) less will be the static position error.

θe =

Static Torque T

Tmax

TL

-θe

0

Angle θ

Version 2 EE IIT, Kharagpur 10 Fig. 9 Representation of the static position error.

Dynamic Response
So far we have discussed about the static performance of step motor. Now we examine how the motor behaves under dynamic condition. Ideally the rotor of a step motor should rotate by step angle once it receives a pulse and the rotation should occur instantaneously. But due to the inertia of the rotor, the rotor cannot settle down at its final position instantaneously and it undergoes through some oscillations as shown in Fig.10. Before settling down after some time. The overshoot increases if the inertia of the rotor is large. This will cause large settling time. In order to decrease the overshoot, additional damping arrangement is sometimes provided. The damping can be in the form of (a) viscous damping, (b) eddy current damping and (c) electronic damping. However the details of the damping arrangements are not discussed in this lesson. θ

t Fig. 10 Dynamic Response of a Step motor θ Start-stop

Slewing t Fig. 11 Start-stop and slewing mode of operation

Continuous Operation
It has been mentioned earlier that though the step motors are normally used for discrete step operations, they can be used for continuous rotation also. A train of control pulses will rotate the motor continuously. Continuous operation can be achieved in two ways: start-stop and slewing mode of operation. In the first case, one pulse is applied, the rotor rotates through one step, settles at its stable location and then another pulse is sent to rotate it by one more step. Thus in each step the rotor comes to a halt before the next pulse is received. On the other hand, the pulses received are at much faster rate in slewing mode and the control circuit generates a pulse before the rotor comes to a steady state. As a result, the rotor runs at uniform speed without stopping after each step. The inertia effect is absent because of the continuous rotation and the motor can take more load when it is slewing. The displacement vs. time characteristics in start-stop and slewing modes are shown in Fig.11. The torque vs. speed characteristics under these two modes have been shown in Fig.12. From this figure it is clear that at a particular speed the torque generated in slewing mode is more in slewing mode. Version 2 EE IIT, Kharagpur 11

Torque

Slewing Start stop Speed Fig. 12 Torque-speed characteristics for continuous running. But there is a problem if we want to start a motor with load under slewing operation. When a motor starts with load, it cannot suddenly start rotating at high speed. It has to be accelerated gradually by increasing the number of steps/sec. This process of starting and acceleration is known as ramping.

Control of Step Motors
In many cases step motors are used for accurate positioning of tools and devices. Precision control over the rotation of the motor is required for these cases. Control of step motors can be achieved in two ways: open loop and closed loop. The open loop control is simpler and more widely used, such a scheme is shown schematically in Fig.13. The command to the pulse generator sets the number of steps for rotation and direction of rotation. The pulse generator correspondingly generates a train of pulse. The Translator is a simple logical device and distributes the position pulse train to the different phases. The amplifier block amplifies this signal and drives current in the corresponding winding. The direction of rotation can also be reversed by sending a direction pulse train to the translator. After receiving a directional pulse the step motor reverses the direction of rotation.
Power Supply Electronic Drive

Position Pulse Train Command Pulse Generator

Translator

Amplifiers

Step Motor

To Load

Direction Pulse Train

Currents to Windings

Fig. 13 Open-loop control of a step motor. The major disadvantage of the open loop scheme is that in case of a missed pulse, there is no way to detect it and correct the switching sequence. A missed pulse may be due to Version 2 EE IIT, Kharagpur 12

malfunctioning of the driver circuit or the pulse generator. This may give rise to erratic behaviour of the rotor. In this sequel the closed loop arrangement has the advantage over open loop control, since it does not allow any pulse to be missed and a pulse is send to the driving circuit after making sure that the motor has rotated in the proper direction by the earlier pulse sent. In order to implement this, we need a feedback mechanism that will detect the rotation in every step and send the information back to the controller. Such an arrangement is shown in Fig. 14. The incremental encoder here is a digital transducer used for measuring the angular displacement.
Motion Commands Control Pulse Train Step Motor Response

Control Computer

Drive System

Measured Pulse Train

Incremental Encoder

Fig. 14 Feedback control of a step motor.

Conclusion
Step motors find wide applications as an electrical actuator. It can be readily interfaced with a computer and can be controlled for rotation at a very small angle, or a precisely determined speed. Step motors nowadays can deliver large torque also. These are the reasons step motors are gradually replacing conventional a.c. and d.c. motors in many cases. The constructions, principles of operation and driving schemes used for step motor have been discussed in this lesson. Both the open loop and close loop control schemes for position control of step motors have been elaborated. The torque characteristics of a step motor is distinctly different from conventional motors and nonlinear in nature. As a result it is not possible even to develop a linearised transfer function model of a step motor (a practise that is common for a.c. and d.c. servomotors). This is possibly a major hindrance for modelling and simulation of systems with step motor drives.

References
1. C.W. de Silva: Control Sensors and Actuators, Prentice Hall, New Jersey, 1989. 2. T. Wildi: Electrical Machine Drives and Power Systems

Review Questions
1. Indicate the correct answer: (a) The major advantage of a permanent magnet step motor is that it can provide holding torque (True/ False). (b) The torque generated in a step motor under start-stop mode is more than under slewing mode (True/ False).

Version 2 EE IIT, Kharagpur 13

(c) A variable reluctance type step motor requires less number of switches than of permanent magnet type (True/ False). (d) Ramping of a step motor refers to slow acceleration during starting (True/ False). (e) 3-phase a.c. excitation is needed to drive a 3-phase step motor (True/ False). 2. A step motor has 130 steps per revolution. Find the input digital pulse rate that produces continuous rotation at a speed of 10.5 revolutions/ sec. 3. The inductance and resistance of each phase winding of a step motor are 10 mH and 2Ω respectively. The switching arrangement has been designed that each phase receives a pulse of duration 10 msec. Is the arrangement satisfactory? Justify your answer. 4. Write down the switching sequence for the step motor shown in Fig. 4 for clockwise rotation with full stepping. 5. Consider the VR type step motor shown in Fig. 5. The switching sequence of the phase winding is 1-3-2-1. Find the direction of rotation and the step angle.

6. What is meant by static position error of a step motor? How can it be reduced? 7. Discuss the limitations of open loop type control of step motors. 8. Can a BLDC motor replace a step motor for increamental motion? What would be the advantages and limitations? 9. Can a step motor replace a BLDC motor for controlled speed drive?

Answer
1. (a) True, (b) False, (c) True, (d) True, (e) False. 2. 1365 pulse/sec. 3. Pulse duration should be at least 30 msec. 5. 150 clockwise direction.

Version 2 EE IIT, Kharagpur 14

Module 7
Electrical Machine Drives
Version 2 EE IIT, Kharagpur 1

Lesson 33
Electrical Actuators: DC Motor Drives
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
After learning the lesson students should be able to A. Describe the major constructional features of dc motors B. Explain the principle of torque generation C. Derive the dynamic speed response characteristics relating armature voltage, load torque and speed D. Describe the realization of a variable voltage controlled source using switch mode power converters. E. Draw the block diagram a typical speed control loop for a separately excited dc motor.

Introduction
Motion control and drives are very important actuation subsystems for Process and Discrete Manufacturing Industries. As we have already seen in Lessons 23, 24 and 31, motion control systems are critical for product quality in discrete manufacturing, while variable speed drives lead to significant energy savings in common industrial loads such as pumps, compressors and fans. Variable speed drives can be categorized into Adjustable Speed Drives and Servo Drives. In adjustable speed drives the speed set points are changed relatively infrequently, in response to changes in process operating pints. Therefore transient response of the drive system is not of consequence. In servo drives, as in CNC machines, set points change constantly (as in contouring systems). While ac motors have replaced dc motors in most of the adjustable speed drive applications. For servo drive applications, dc motors are still used, although they are also being replaced by BLDC motors. In this lesson we discuss speed and position control with dc motors. The next lesson discusses adjustable speed drives using induction motors, while Lesson 35 discusses BLDC servo drives.

DC Servomotors
Direct current servomotors are used as feed actuators in many machine tool industries. These motors are generally of the permanent magnet (PM) type in which the stator magnetic flux remains essentially constant at all levels of the armature current and the speed-torque relationship is linear. Direct current servomotors have a high peak torque for quick accelerations. A cross-sectional view of a typical permanent magnet dc servomotor is shown in Fig. 33.1.

Mechanical Construction
Stator consists of Yoke and Poles and provides mechanical support to the machine. The yoke provides a highly permeable path for magnetic flux. It is made of cast steel. Field poles are made of thin laminations stacked together. This is done to minimize the magnetic losses due to Version 2 EE IIT, Kharagpur 3

the armature flux. The cross sectional area of the field pole is less than that of the pole shoe. The pole shoe helps to establish a uniform flux density around the air gap. Field winding: DC excitations are provided to field windings wound on pole shoes to create electromagnetic poles of alternating polarity. Depending on the connections of field windings DC motors may be termed as shunt, series, compound or separately excited. Shunt motors have field winding connected in parallel with the armature winding while series motors have the field winding connected in series with the armature winding. A compound dc machine may have both field windings wound on the same pole. Smaller DC servomotors generally have permanent magnets for poles. Armature – The rotating part of a dc machine is called the armature. The length of the armature is usually the same as that of the pole. It is made of thin, highly permeable, and electrically insulated circular steel laminations that are stacked together and rigidly mounted on the shaft. The laminations have axial slots on their periphery to house the armature coils. Insulated copper wires are typically used for the armature coils to achieve a low armature resistance. Commutator – The commutator is made of wedge – shaped hard-drawn copper segments. Sheets of mica insulate the copper segments from one another. One end of the armature coil is electrically connected to a copper segment of the commutator. The commutators rotate with the armature keeping a sliding contact with the brushes, which remain stationary. Brushes: Brushes are held in a fixed position by means of brush holders and remain in sliding contact with the commutator segments. An adjustable spring inside the brush holder exerts a constant pressure on the brush in order to maintain a proper contact between the brush and the commutator. The brushes are connected to the armature terminals of the machine. The material for the brush is normally carbon or carbon-graphite.

Version 2 EE IIT, Kharagpur 4

Tangential magnet

Feedback yoke Radial magnet Pole sheet

Fig. 33.1 Cross-section of a permanent magnet-excited dc servomotor. Flux lines Field winding Air gap N Frame or yoke

S Pole core

• • •
Pole shoe

• • •
Shaft θr

• •

• • •

S

N

Laminated armature core

Field pole Fig. 33.2 Diagrammatic sketch of a D.C. machine. Version 2 EE IIT, Kharagpur 5

Principle of Operation
The cross-sectional view of a DC motor has been shown in Fig. 33.2. Consider a particular position in space between stator and rotor. Whichever conductor is present there, will have current flowing through it, which depends on the applied armature voltage. This current would produce a flux which would interact with the field flux to produce torque. In course of rotation of the armature adjacent conductors will occupy this position in space. No matter which conductor comes to that particular position at any given point of time, it will have same current flowing through it. This is true for all the positions although the magnitude and polarity of the torque produced by individual conductors in different positions may be different. The polarity of the torque is identical for conductor positions under north or south pole, since the direction of the current flowing through it at that position is unique, given the direction of rotation and the applied armature voltage due to the commutators slipping over the brushes, as shown in Fig. 33.3. N -a Rotation Carbon brush Copper commutator segments.

+ a S

Fig. 33.3 Brush and commutator positions in a DC motor

Points to Ponder: 1
A. Why is it that dc motors are preferred for control applications, such as actuation, but ac motors are preferred for high power applications, such as compressors and fans ? B. In a dc motor, is the field flux stationary or rotating? Is the armature flux stationary or rotating?

Version 2 EE IIT, Kharagpur 6

To quantify the net torque produced,

N

F

D

S

Fig. 33.4 Principle of rotation.
πD where D is the diameter of the armature and 2p p be number of pole pairs. Let L be the length of the pole = length of the conductor, then pole πD area = YL = L 2p Φ 2pΦ ∴Flux density B = Φ/(YL) = . = πD πDL L 2p Force (f) experienced by an armature conductor carrying current Ic =B.Ic.L: Therefore, torque D 2pΦ D 1 experienced by the conductor = B Ic L = .I c .L. = I c .p.Φ . 2 πDL 2 π

let Φ be the flux per pole and pole pitch be Y =

The principle has been demonstrated in Fig. 33.4. If the armature current is Ia, then conductor I current Ic = a where c is the number of parallel paths. If z be the total number of conductors, c 1 I then total torque developed T = . a .z.p.Φ Nm, if Ia is in Ampere and Φ is in Wb. π c I 1I T = a .z.p.Φ = 0.318 a zpΦ Nm. π c c One can therefore see that the torque produced is proportional to the armature current, if the flux can be assumed to be constant.

Version 2 EE IIT, Kharagpur 7

Ia + Ra φ If + Vf n T P Fig. 33.5 Schematic of a separately excited DC motor

+

V

E -

Two equations are required to define the behavior of a dc machine: the torque and the voltage equations. Fig. 33.5 describes a schematic of a separately excited DC motor. Where T = magnetic torque, N.m φ = flux per pole, Wb /a = current in armature circuit, A E = induced voltage (emf), V φ = angular velocity, rad/s Kf = constant determined by design of winding If = Field current n = speed of the motor in rpm Vf = field voltage P = mechanical power The torque equation relates the torque, to the armature current: T = K f φ Ia (33.1) and the voltage equation relates the induced voltage in the armature winding to the rotational speed: E = K f φω (33.2) For a motor, an input voltage V is supplied to the armature, and the corresponding voltage equation becomes E = V − I a Ra = K f φω (33.3) where Ra is the resistance of the armature circuit and IaRa is the voltage drop across this resistance. The armature inductance is negligible in Eq.(33.1). Equation 33.3 multiplied by Armature current Ia, yields the power equation, 2 P = ωT = VI a − I a Ra (33.4) 2 where P is the mechanical output power, VIa is the electrical input power, and I a Ra is the Version 2 EE IIT, Kharagpur 8

electrical power loss.

T = Kt I a (33.5) V − I a Ra = Kvω (33.6) The parameters Kt and Kv are referred to as the torque and voltage constants. In SI units the torque constant in Newton-meters per ampere equals the voltage constant in volt-seconds per radian. Zones of steady state torque-speed characteristics of the motor are shown in Fig. 33.6. Note that constant torque characteristics can be maintained by armature voltage control up to a certain speed. At the rated speed this would require rated voltage to be applied to the armature voltage, and further increase would not be possible due to limitations of the motor such as insulation ratings and thermal ratings. Speed increase beyond this point would only be possible at the cost of a reduction in torque and the machine will operate in the constant power mode. The corresponding power characteristics are shown in dotted lines.
Power, P

0

Torque, T Speed, ω Armature current, Ia Field current, If

Speed, ω Constant Constant torque power Fig. 33.6 Speed, torque and power characteristics of separately excited DC motors 0

Modern dc motors often use a permanent-magnet (PM) field, rather than an externally excited field. Both types are referred to as dc servomotors and are characterized by Eqs. (33.5) and (33.6). The PM obviates the need for a field voltage source and results in higher efficiency and fewer thermal problems. The dc servomotor drives a mechanical load consisting of dynamic and static components: dω T=J + Ts ………………………………...(33.7) dt Where J is the combined moment of inertia of the motor and load, and Ts is the static load due to friction and cutting forces in NC systems. Elimination of Ia and T from Eqs. (33.5) through (33.7), and rearrangement of the terms so as to separate the independent variables, gives the speed equation R 1 dω τm +ω = V − a Ts …………………………(33.8) dt Kv Kt K v where τm is the mechanical time constant of the loaded motor and is defined by JRa τm = ……………………………………(33.9) Kt K v The Laplace transform of Eq. (33.8) is Version 2 EE IIT, Kharagpur 9

K mV ( s ) − ( Ra K m / K t ) Ts ( s ) ………………..(33.10) 1 + sτ m where Km is the gain of the motor and is defined by Km = 1/KV.

ω (s) =

Points to Ponder: 2
A. Can you identify some of the assumptions that have been made in the derivation of the above model? B. What can you say about the input-output transfer function of the dc motor?

Braking methods in servo-drive
Braking of a motor is a normal requirement of many industrial applications, such as CNC machines to stop the slide/spindle to the programmed position or within a definite distance in case of power failure or emergency conditions. There are two types of braking employed in servo-drives.
Ia A Ra Lf V E Vf Rb E F If Ia A1 Ra Lf Vf F If

+M
A2 F2

Rf

+M
A2 F2

Rf

Fig. 33.7 (a) Regenerative braking (b) Dynamic braking Dynamic braking (Fig. 33.7(b)) Braking is realized by shorting the armature leads through contactor and dissipating the kinetic energy stored in the motor into the Dynamic Braking Resistor (DBR) in the form of heat. During this period, reverse torque will be generated which will bring the motor to a stand still faster. This type of braking is a fail safe braking and finds application particularly during mains failure and emergency situations. Regenerative braking (Fig. 33.7(a)): Regenerative braking is possible if the motor is driven by the stored mechanical energy of the load and energy is returned to the source, i.e., dc link or the mains. The feeding of power back to dc source raises the dc link voltage. Depending on the load conditions and speed, this can reach dangerous levels unless the additional energy is returned to ac mains by using the converter in the inverter mode. Thus, regenerative braking is possible only with fully controlled drives.

Version 2 EE IIT, Kharagpur 10

Transistor PWM dc Converter
Transistor dc drives are ideal for controlling dc servomotors. The transistors are commonly used in the switching mode at frequencies between 1 kHz and 10 kHz (pulse width modulated). In the PWM technique the average dc voltage is proportional to the pulse width.

L1 L2 L3

VC

+ -

iC C

Q1 Q2

Fig.

Ia La Ra V

E

Q4 Q3

D4 D3

D2

Fig. 33.8 Transistor dc four quadrant amplifier.

Figure 33.9 shows the four-quadrant operation of a dc drive. Four-quadrant operation of a drive is enables it for: (a) Forward running—quadrant I (b) Reverse running—quadrant III (c) Forward braking—quadrant II (d) Reverse braking—quadrant IV Fig. 33.8 shows the basic diagram of a transistor dc four-quadrant amplifier. A rectifier is fed from the three phase ac line and delivers power into a dc bus. The buffer capacitor C can supply stored energy for acceleration and can accept energy as long as the motor absorbs mechanical energy during braking. The buffer capacitor is thus working as generator and supplies electrical energy. The capacitor is so chosen that the dc bus voltage changes only slightly. The motor can be controlled selectively for clockwise or counterclockwise rotation and can be accelerated or braked by controlling two diagonally opposite transistors (Q1 - Q3 or Q2 - Q4). The magnitude of the armature voltage V and thus the speed n is determined by pulse width modulation of transistors acting as switches for a switched transistor chopper. Two energy storages are necessary to operate a transistor controller in all four quadrants:

• •

a large capacitor C which maintain the voltage VC constant and is capable of accepting energy to store and to deliver. an inductance, which smoothens the motor current and acts as an energy storage element. This is especially important during braking mode. At high switching frequencies of the drive, the armature inductance La of the motor is generally sufficient.

Version 2 EE IIT, Kharagpur 11

Speed +n Forward braking Forward driving

Torque

Regeneration inverting -T Reverse driving

II III

I IV

Motoring rectifying Reverse braking +T

Torque

Motoring rectifying -n Speed

Regeneration inverting

Fig. 33.9 Four-quadrant operation of a dc drive

Driving, clockwise (CW), I quadrant
The voltage and the current waveforms for this mode are shown in Fig. 33.10. Assume that Logic ‘0’ represents that a transistor is ON. The following sequence of phases occurs.

From the figure, at the time t0, both transistors Ql and Q3 are ‘ON’. The armature voltage of the motor is positive, i.e., V = VC and current flows through the motor via Ql, La, Ra and Q3. At time t1, switch Ql is opened. Instantaneously, due to the armature inductance, the motor current commutates from Q1 to diode D2, so that current through the inductance does not reverse. This current is not flowing through the dc bus any more, but is circulated by the stored energy in the armature inductance against the back emf of the motor. in the lower half of the bridge from Q3 through D2, La, Ra, motor and back to Q3 (free wheeling), at this point the motor voltage V and the dc bus current Ia are zero. At time t2, the situation is the same as it was at time t1. At time t3, switch Q3 is open. The motor current now commutes through diode D4 and circulates in the upper half of the bridge circuit via Q1, La, Ra, motor D4 and Q1, The motor voltage and dc link current go to zero again. At time t4, switch Q3 is closed and a new switching cycle begins.

The mean value of the motor voltage V depends on the ratio between the switch ‘ON’ time tON and the switch OFF time tOFF. During the switch ON time tON the energy is derived from the dc bus while during the period tOFF, assuming the current does not go to zero, the current is driven by the energy stored in the inductance. The motor thus maintains a positive product from voltage V and current Ia during both the periods tON and tOFF and thus converts electrical energy into mechanical energy. For simplicity the drop across the motor (Ia Ra) is neglected since it is small compared to the induced voltage E in the motor. Driving in the CCW direction is similar. Version 2 EE IIT, Kharagpur 12

t0 Q1 Q2 Q3 Q4 V

t1 t2

t3 t4 Q1 Q2 Q3 Q4 V E V

t0

t1 t2

t3 t4

tE Ia

tA tE Ia Ia tA

V E

iZ Driving clockwise t0 Q1 Q2 Q3 Q4 V V E t1 t2 t3 t4

iZ

Ia

Driving counter-clockwise t0 Q1 Q2 Q3 Q4 V V E Ia Ia t1 t2 t3 t4

Braking clockwise

Braking counter-clockwise

Fig. 33.10 Voltage and the current wave forms for Fomquadrant Drive.

Version 2 EE IIT, Kharagpur 13

Braking, clockwise, IV quadrant
For rapid braking, the torque has to reverse, while the speed reduces but the motor continues in the same direction. Thus, the mean value of the motor current IA and the armature voltage V must have opposite signs for a back flow of actual power out of the motor circuit. For stopping the motor, the mean value of the armature voltage V is reduced compared to voltage induced in the motor E. Thus, after the magnetic energy stored in the motor inductance is exhausted, The direction of current within the motor circuit is reversed. The transition from driving to braking is the result of decrease in tON and an increase in tOFF. At time tl, switch Q2 is closed. The voltage E drives a current through Ra, La, Q2 and D3. The energy is stored in the La inductance which is given by 1 WLA = I a L2a 2 where WLA = energy stored La = inductance and Ia = current through the motor.

At time t2, switch Q2 is OFF and current Ia commutates over to diode Dl and flows into the dc bus charging the capacitor and then returns through D3 to the motor. The voltage induced in the inductance and the induced motor voltage is in series. Their sum is larger than the voltages delivered from the dc bus. Energy is thus fed back into dc bus and stored in the capacitor as given by 1 2 WC = I A CVC 2 where WC = energy stored C = capacitance and VC = dc link voltage Thus, the voltage of the dc link increases. If now at time t3, the switch Q4 is turned ON, the armature current will flow in the upper circuit through Ra, La, Dl and Q4 and energy will again be stored in La. This energy in turn is fed into the dc bus at time t4 via diodes D1 and D3. This cycle is repeated periodically. The time period t4 is for storing the energy in the inductance and period tON is for feeding the energy back into dc bus. The mean values of the motor current and motor voltage have opposite polarity. The motor is braked with mean constant torque because the actual power is fed back to the source. Again, the cycle of operation during CCW braking is similar.

Advantages of transistor PWM dc drives over thyristor drives
(a) PWM transistor drive has a high form factor of approximately 1. (b) Less heating of the motor and an increased torque output (about 20% more than thyristor drives). (c) High banwidth dynamic response resulting in better surface finish, low machining stress and no resonance problems for machine tools. (d) Increased device reliability of the transistor.

Version 2 EE IIT, Kharagpur 14

Therefore transistor PWM amplifiers are used advantageously as servo drives for fast and accurate actuation of industrial machines.

Points to Ponder: 3
A. Can you describe what happens in dynamic braking? B. There are drives that are cheaper and permit restricted operations, such as operation only in 2 quadrants or even a single quadrant. Study the operation if the dc motor is fed from a single phase fully controlled rectifier. Compare the resource requirements of this configuration with the one presented above. Having presented the basic principles of armature voltage control using a controlled variable voltage source such as the PWM switching converter, below we describe the over all control loops for speed control.

Closed loop of control of DC motors
Vr(s) Ve + Vb K2 Va + E(s) KB K1 Fig. 33.11 Block diagram of closed-loop speed control of separately excited DC motor
1 R a (sτ a + 1)

Ia KT

Td +

TL(s) -

ω(s) 1 B L (sτ L + 1)

The basic structure of the control loop is shown in Fig. 33.11. For servo drives, the speed control loop is an inner loop of the cascade structure and is generally proportional in nature. In independent speed control applications this may be designed as a PI loop also. Speed may be fed back from a tachogenerator, or derived from position measurements by differentiation. A practical implementation structure of the above control loop is shown below in Fig. 33.12. This typically includes a current control loop for torque control as well as a field voltage control loop. Note the PI speed and current control loops. Additional circuitry needed for practical implementations, such a filters for noise removal in current and speed feedback channels as well as current limiters in set points to avoid phenomena such as integrator windup.

Version 2 EE IIT, Kharagpur 15

ac supply ωref + ωm Ia(ref)
PI Speed controller Current limiter Current Vc controller Firing α circuit

+

Ia Ra

Ia Ia
Filter

-

Va
+

Va ac supply αf Ir M

E(ref) +

E
Field controller

Vcf

Firing circuit

Filter

Techno generator Fig. 33.12 Schematic of practical implemented of closed loop control for separately excited DC motor.

Points to Ponder: 3
A. Draw the control loop structure for a position control application B. Explain why the need for the filters and the current limiter in the block diagram of Fig. 33.12.

Lesson summary
In this lesson, the following topics related to CNC machines have been discussed. A. Reference pulse and reference word interpolators B. Linear and Circular Interpolation C. Digital Integration with a DDA D. Open loop and closed loop control E. Control of PTP and Contouring Systems F. Characteristics of Feed and Spindle Drives

Version 2 EE IIT, Kharagpur 16

Points to Ponder: 1
A. Why is it that dc motors are preferred for control applications, such as actuation, but ac motors are preferred for high power applications, such as compressors and fans ?
Ans: Dc motors have simpler control and faster dynamic response, which is why it is preferred for servo applications. AC motors, on the other hand have better size to power ratio and lesser maintenance problems than DC motors, which is why they are preferred for large power applications.

B. In a dc motor, is the field flux stationary or rotating? Is the armature flux stationary or rotating?
Ans: In a DC motors both the field and armature fluxes are stationary in space, although the armature is rotating.

Points to Ponder: 2
A. Can you identify some of the assumptions that have been made in the derivation of the above model?
Ans: There are several assumptions. The first one is that of linearity which implies, among other things, that there is no flux saturation. Many factors are neglected, such armature inductance, armature reaction, brush drops. Friction is assumed viscous too. Many other assumptions can be found.

B. What can you say about the input-output transfer function of the dc motor?
Ans: The transfer function is first order. This is the result of neglecting the armature electrical time constant in comparison with the mechanical time constant. The transfer function would be second order too for a position control application, since the integrator between speed and position would now be included.

Points to Ponder: 3
A. There are drives that are cheaper and permit restricted operations, such as operation only in 2 quadrants or even a single quadrant. What are the quadrants of operation if the dc motor is fed from a single phase fully controlled thyristor rectifier. Compare the resource requirements of this configuration with the one presented above.
Ans: A single phase fully controlled rectifier can operate only in Quadrant I and IV. This is because, with such drives, while voltage can be reversed by controlling the firing angle, current cannot be reversed. Note that such drives can only be used for forward motoring, but not forward braking. The quadrant IV operation of regenerative braking can be used with loads that can drive the load in opposite directions, such as overhauling loads. For four-quadrant operation, two such drives must be connected in antiparallel.

Version 2 EE IIT, Kharagpur 17

Note that the number of devices required is also halved. For details see any standard text on Electric Drives. B. Can you describe what happens in dynamic braking?
Ans: In dynamic braking the armature is connected across a braking resistor. The current continues in the same direction till the magnetic energy stored in the armature inductance is spent. Then, the back emf drives current in the reverse direction and mechanical power is spent as heat in the resistor, thus braking the motor.

Points to Ponder: 4
A. Draw the control loop structure for a position control application.
Vr(s) Ve K4 + Vb K2 Va + E(s) KB K1 K3 Fig. 33.13 Block diagram of closed-loop position control of separately excited DC motor. Ia KT TL(s) -

θr + -

1 R a (sτ a + 1)

Td +

ω(s) 1 B L (sτ L + 1)

1

θ(s)
S

B. Explain why the need for the filters and the current limiter in the block diagram of Fig. 33.12.
Ans: The current limiter is needed to avoid integrator windup in the current loop as well as to prevent peak current surges in the motor for large speed errors, while maintaining a high loop gain in steady state operations. The filters are needed to prevent sensor noise from causing noisy inputs into the motor.

Version 2 EE IIT, Kharagpur 18

Module 7
Electrical Machine Drives
Version 2 EE IIT, Kharagpur 1

Lesson 34
Electrical Actuators: Induction Motor Drives
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
After learning the lesson students should be able to understand A. Concept of slip B. Equivalent circuit of induction motor. C. Torque-speed characteristics. D. Methods of induction motor speed control. E. Principles of PWM inverter. F. Implementation of constant V/f control.

Introduction
For adjustible speed applications, the induction machine, particularly the cage rotor type, is most commonly used in industry. These machines are very cheap and rugged, and are available from fractional horsepower to multi-megawatt capacity, both in single-phase and poly-phase versions. In this lesson, the basic fundamentals of construction, operation and speed control for induction motors are presented. In cage rotor type induction motors the rotor has a squirrel cage-like structure with shorted end rings. The stator has a three-phase winding, and embedded in slots distributed sinusoidally. It can be shown that a sinusoidal three-phase balanced set of ac voltages applied to the three-phase stator windings creates a magnetic field rotating at angular speed ωs = 4πfs /P where fs is the supply frequency in Hz and P is the number of stator poles. If the rotor is rotating at an angular speed ωr , i.e. at an angular speed (ωs - ωr) with respect to the rotating stator mmf, its conductors will be subjected to a sweeping magnetic field, inducing voltages and current and mmf in the short-circuited rotor bars at a frequency (ωs - ωr)P/4π, known as the slip speed. The interaction of air gap flux and rotor mmf produces torque. The per unit slip ω is defined as ω − ωr S= s ωs

Equivalent Circuit
Figure 34.1 shows the equivalent circuit with respect to the stator, where Ir is given as Vm Ir == ⎛ Rr ⎞ ⎜ ⎟ + jωe Llr ⎝ S ⎠ and parameters Rr and Llr stand for the resistance and inductance parameters referred to to the stator. Since the output power is the product of developed electrical torque Te and speed ωm, Te can be expressed as

Version 2 EE IIT, Kharagpur 3

⎛P⎞ R Te = 3 ⎜ ⎟ I r2 r ⎝ 2 ⎠ Sωe
RS lr VS Lm
Rr S

Lls

Llr

Fig. 34.1 Approximate per phase equivalent circuit

In Figure 34.1, the magnitude of the rotor current Ir can be written as
Ir = Vs
2 ( R s + R r S) + ω e2 ( L ls + Llr ) 2

This yields that,
Vs2 ⎛P⎞ R Te = 3 ⎜ ⎟ r ⋅ 2 2 ⎝ 2 ⎠ Sω e ( R s + R r S ) + ω e2 ( L ls + L lr )

Torque-Speed Curve
The torque Te can be calculated as a function of slip S from the equation 1. Figure 34.2 shows the torque-speed (ω r / ω e = 1 − S ) curve. The various operating zones in the figure can be defined as plugging (1.0 < S < 2.0), motoring (0 < S < 1.0), and regenerating (S< 0). In the normal motoring region, Te = 0 at S = 0, and as S increases (i.e., speed decreases), Te increases in a quasi-linear curve until breakdown, or maximum torque Tem is reached. Beyond this point, Te decreases with the increase in S.

Version 2 EE IIT, Kharagpur 4

Plugging Torque (Tem)

Motoring Tem

Regenerating Maximum or breakdown torque

Torque

Tes 2 1 Starting torque 0 ⎛ω Speed ⎜ r ⎝ ωe ⎞ ⎟ pu ⎠ Slip(S) 0 1

Synchronous speed

Teg Fig. 34.2 Torque-speed curve of induction motor

In the regenerating region, as the name indicates, the machine acts as a generator. The rotor moves at supersynchronous speed in the same direction as that of the air gap flux so that the slip becomes negative, creating negative, or regenerating torque (Teg). With a variable-frequency power supply, the machine stator frequency can be controlled to be lower than the rotor speed (ωe < ωr) to obtain a regenerative braking effect.

Speed Control
From the torque speed characteristics in Fig. 34.2, it can be seen that at any rotor speed the magnitude and/or frequency of the supply voltage can be controlled for obtaining a desired torque. The three possible modes of speed control are discussed below.

Variable-Voltage, Constant-Frequency Operation
A simple method of controlling speed in a cage-type induction motor is by varying the stator voltage at constant supply frequency. Stator voltage control is also used for “soft start” to limit the stator current during periods of low rotor speeds. Figure 34.3 shows the torque-speed curves with variable stator voltage. Often, low-power motor drives use this type of speed control due to the simplicity of the drive circuit.

Variable-Frequency Operation
Figure 34.4 shows the torque-speed curve, if the stator supply frequency is increased with constant supply voltage, where ωe is the base angular speed. Note, however, that beyond the rated frequency ωb , there is fall in maximum torque developed, while the speed rises.

Version 2 EE IIT, Kharagpur 5

1.0

Speed control range 100% stator 1.0 V s voltage 0.75 Vs

0.75 ⎛ T Torque ⎜ e ⎝ Tem ⎞ ⎟ pu ⎠ 0.50

0.50 Vs 0.25 0 0.25 Vs 1.0

0.2

0.4

0.6

0.8

⎛ω ⎞ Speed ⎜ r ⎟ pu ⎝ ωe ⎠ Fig. 34.3 Torque-speed curves at variable supply voltage

1.0

Tem Rated curve
2 Tem ωe = constant

⎛ T Torque ⎜ e ⎝ Tem

⎞ ⎟ pu ⎠

0.50

0

1

⎛ω Frequency ⎜ e ⎝ ωb

2 ⎞ ⎟ pu ⎠

3

Fig. 34.4 Torque-speed curves at variable stator frequency

Variable voltage variable frequency operation with constant V/f
Figure 34.5 shows the torque-speed curves for constant V/f operation. Note that the maximum torque Tem remains approximately constant. Since the air gap flux of the machine is kept at the rated value, the torque per ampere is high. Therefore fast variations in acceleration can be achieved by stator current control. Since the supply frequency is lowered at low speeds, the machine operates at low slip always, so the energy efficiency does not suffer.

Version 2 EE IIT, Kharagpur 6

Vs Maximum torque ω = constant e

1.0

Rated curve

⎛ T Torque ⎜ e ⎝ Tem

0.50 ⎞ pu ⎟ ⎠ 0

⎛ω Frequency ⎜ e ⎝ ωb

0.5

⎞ ⎟ pu ⎠

1.0

Fig. 34.5 Torque-speed curves at constant V/f

Majority of industrial variable-speed ac drives operate with a variable voltage variable frequency power supply.

Points to Ponder: 1
A. In what type of applications, would it make sense to prefer a simple stator voltage control, rather than a constant V/f control ? B. How can you be in the plugging region of the torque speed curve shown in Fig. 34.2 ?

Variable Voltage Variable Frequency Supply

Three-phase supply

Cf

+

Vdc

Induction Motor

Diode dc link filter PWM Rectifier Inverter Fig. 34.6 PWM inverter fed induction motor drive

The variable voltage variable frequency supply for an induction motor drive consists of a uncontrolled (Fig. 34.6) or controlled rectifier (Fig. 34.7) (fixed voltage fixed frequency ac to variable/fixed voltage dc) and an inverter (dc to variable voltage/variable frequency ac). If rectification is uncontrolled, as in diode rectifiers, the voltage and frequency can both be controlled in a pulse-width-modulated (PWM) inverter as shown in Figure 34.6. The dc link filter consists of a capacitor to keep the input voltage to the inverter stable and ripple-free.

Version 2 EE IIT, Kharagpur 7

Three-phase ac power supply

Cf

+

Vdc

Induction Motor

Controlled dc link filter PWM Rectifier Inverter Fig. 34.7 Variable-voltage, variable-frequency (VVVF) induction motor drive

On the other hand, a controlled rectifier can be used to vary the dc link voltage, while a square wave inverter can be used to change the frequency. This configuration is shown in Fig. 34.7.
Controlled Rectifier -II (regeneration)

Three-phase ac power supply

Cf V Controlled Rectifier -I (motoring)

+

Vdc

Motor

dc link filter

PWM Inverter

Induction or Synchronous

Fig. 34.8 Regenerative voltage-source inverter-fed ac drive.

To recover the regenerative energy in the dc link, an antiparallel-controlled rectifier is required to handle the regenerative energy, as shown in Fig. 34.8. The above are basically controlled voltage sources. These can however be operated as controlled current sources by incorporating an outer current feedback loop as shown in Fig. 34.9.

Version 2 EE IIT, Kharagpur 8

Three-phase ac power supply

V Rectifier
dc link filter

Cf

+

Vdc

PWM Inverter - ias

Induction Motor

+

i as i
∗ bs

-

ib + ic +

i cs

Fig. 34.9 Current-controlled voltage-source-driven induction motor drive

Points to Ponder: 2
A. Why is it that drive with the single controlled rectifier shown in Fig. 34.7 cannot be used for regenerative braking? B. How does one generate a controlled current source out of a voltage source inverter?

Voltage-source Inverter-driven Induction Motor
A three-phase variable frequency inverter supplying an induction motor is shown in Figure 34.10. The power devices are assumed to be ideal switches. There are two major types of switching schemes for the inverters, namely, square wave switching and PWM switching.

Square wave inverters
The gating signals and the resulting line voltages for square wave switching are shown in Figure 34.11. The phase voltages are derived from the line voltages assuming a balanced three-phase system.

Version 2 EE IIT, Kharagpur 9

T1 Vdc + T4

D1 a

T3

D3 b

T5

D5 c

D4

T6

D6

T2

D2

Induction Motor

Fig. 34.10 A schematic of the generic inverter-fed induction motor drive.

Version 2 EE IIT, Kharagpur 10

G1 G2 G3 G4 G5 G6 0 0 0 0 0 0 Vdc Vab 0 - Vdc 0 60º 120º 180º 240º 300º 360º

Vbc

Vca 0 2Vdc/3 Vdc/3 Vas 0 -Vdc/3 -2Vdc/3

Vbs

0

Vcs

0

Fig. 34.11 Inverter gate (base) signals and line-and phase-voltage waveforms

The square wave inverter control is simple and the switching frequency and consequently, switching losses are low. However, significant energies of the lower order harmonics and large distortions in current wave require bulky low-pass filters. Moreover, this scheme can only Version 2 EE IIT, Kharagpur 11

achieve frequency control. For voltage control a controlled rectifier is needed, which offsets some of the cost advantages of the simple inverter.

PWM Principle
It is possible to control the output voltage and frequency of the PWM inverter simultaneously, as well as optimize the harmonics by performing multiple switching within the inverter major cycle which determines frequency. For example, the fundamental voltage for a square wave has the maximum amplitude (4Vd/π) but by intermediate switching, as shown in Fig. 34.12, the magnitude can be reduced. This determines the principle of simultaneous voltage control by PWM. Different possible strategies for PWM switching exist. They have different harmonic contents. In the following only a sinusoidal PWM is discussed.
V1

Vao

+Vd

0

π

Fig. 34.12 PWM principle to control output voltage.

Sinusoidal PWM
Figure 34.13(a) explains the general principle of SPWM, where an isosceles triangle carrier wave of frequency fc is compared with the sinusoidal modulating wave of fundamental frequency f, and the points of intersection determine the switching points of power devices. For example, for phase-a, voltage (Va0) is obtained by switching ON Q1 and Q4 of half-bridge inverter, as shown in the figure 13. Assuming that f << fc , the pulse widths of va0 wave vary in a sinusoidal manner. Thus, the fundamental frequency is controlled by varying f and its amplitude is proportional to the command modulating voltage. The Fourier analysis of the va0 wave can be shown to be of the form: v a0 = 0.5mVd sin ( 2πft + φ ) + harmonic frequency terms

Version 2 EE IIT, Kharagpur 12

VT

vp

vao*

vbo*

vco* Carrier wave

0

t Sinewave signal

vao

+0.5Vd Vd t -0.5Vd

0

Fig. 34.13(a) Principle of sinusoidal PWM for three-phase bridge inverter.
+Vd Vab 0 ωt

Fig. 34.13(b) Line voltage waves of PWM inverter

where m = modulation index and φ = phase shift of output, depending on the position of the modulating wave. The modulation index m is defined as V m= P VT where Vp = peak value of the modulating wave and VT = peak value of the carrier wave. Ideally, m can be varied between 0 and 1 to give a linear relation between the modulating and output wave. The inverter basically acts as a linear amplifier. The line voltage waveform is shown in Fig. 34.13(b).

Implementation of a constant voltage/constant frequency strategy
An implementation of the constant volts/Hz control strategy for the inverter-fed induction motor in close loop is shown in Figure 34.14. The frequency command f s* is enforced in the inverter and the corresponding dc link voltage is controlled through the front-end converter.

Version 2 EE IIT, Kharagpur 13

Three-phase ac power supply

Controlled Rectifier Vdc

+ -

Cf

VVVF Inverter

Induction Motor

Kdc V0 + + Kvf fs
*

Tach

1 2π
ω
* r

+ -

PI controller

Te

*

Limiter

ωsl + + ωr

*

Fig. 34.14 Closed-loop induction motor drive with constant volts/Hz control strategy.

An outer speed PI control loop in the induction motor drive, shown in Figure 34.14 computes the frequency and voltage set points for the inverter and the converter respectively. The limiter ensures that the slip-speed command is within the maximum allowable slip speed of the induction motor. The slip-speed command is added to electrical rotor speed to obtain the stator frequency command. Thereafter, the stator frequency command is processed in an open-loop drive. Kdc is the constant of proportionality between the dc load voltage and the stator frequency.

Points to Ponder: 3
A. Explain how the scheme in Fig. 34.10 achieves constant V/f. B. Name two advantages and two disadvantages of a PWM inverter over a square wave inverter.

Lesson Summary
In this lesson, the following topics related to Induction Motor Drives have been discussed. a. b. c. d. e. f. Concept of slip Equivalent circuit of induction motor. Torque-speed characteristics. Methods of induction motor speed control. Principles of PWM inverter. Implementation of constant V/f control.

Version 2 EE IIT, Kharagpur 14

Answers, Remarks and Hints to Points to Ponder Points to Ponder: 1
A. In what type of applications, would it make sense to prefer a simple stator voltage control, rather than a constant V/f control?
Ans: For very simple applications with low torque demands and/or speed ranges a stator voltage control scheme would be adequate. Domestic fan controllers are of this type. Note that, for such applications, the cost of the controller is much more important than the improvement in efficiency, from a commercial point of view. The principle of stator voltage control is also used for soft starting of motors.

B. How can you be in the plugging region of the torque speed curve shown in Fig. 34.2?
Ans: Suppose the motor is rotating in the forward direction with a slip s. If suddenly the phase sequence of the three-phase supply is reversed, the slip would be equal to (1+(1-s)), or 2-s. Thus the motor would be in the plugging region of the torque speed curve.

Points to Ponder: 2
A. Why is it that drives with the single controlled rectifier shown in Fig. 34.7 cannot be used for regenerative braking in the forward direction?
Ans: Because the current through the controlled rectifiers cannot be reversed. One thus needs another rectifier connected in anti-parallel.

B. How does one generate a controlled current source out of a voltage source inverter?
Ans: By feedback control of the voltage input to the inverter. Thus the duty ratio of the PWM control input, which basically changes the voltage and thereby the current, is manipulated to make a current equal to the set point flow.

Points to Ponder: 3
A. Explain how the scheme in Fig. 34.10 achieves constant V/f.
Ans: Note that the voltage setpoint to the rectifier and the frequency setpoint to the inverter are related by a constant. Thus constant V/f is maintained.

B. Name two advantages and two disadvantages of a PWM inverter over a square wave inverter.
Ans: Advantages are simultaneous voltage and frequency control, and control over harmonic content. Disadvantages are complexity of switching law and requirement of fast switching.

Version 2 EE IIT, Kharagpur 15

Module 7
Electrical Machine Drives
Version 2 EE IIT, Kharagpur 1

Lesson 35
Electrical Actuators: BLDC Motor Drives
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
After learning the lesson students should be able to A. Define the Structure of a PM BLDC Motor. B. Describe the principle of operation of a PM BLDC motor. C. Understand Closed Loop Control of a BLDC Drive. D. Name applications of BLDC Motor.

Introduction
Brushless DC motors, rather surprisingly, is a kind of permanent magnate synchronous motor. Permanent magnet synchronous motors are classified on the basis of the wave shape of their induce emf, i.e, sinusoidal and trapezoidal. The sinusoidal type is known as permanent magnet synchronous motor; the trapezoidal type goes under the name of PM Brushless dc (BLDC) machine. Permanent magnet (PM) DC brushed and brushless motors incorporate a combination of PM and electromagnetic fields to produce torque (or force) resulting in motion. This is done in the DC motor by a PM stator and a wound armature or rotor. Current in the DC motor is automatically switched to different windings by means of a commutator and brushes to create continuous motion. In a brushless motor, the rotor incorporates the magnets, and the stator contains the windings. As the name suggests brushes are absent and hence in this case, commutation is implemented electronically with a drive amplifier that uses semiconductor switches to change current in the windings based on rotor position feedback. In this respect, the BLDC motor is equivalent to a reversed DC commutator motor, in which the magnet rotates while the conductors remain stationary. Therefore, BLDC motors often incorporate either internal or external position sensors to sense the actual rotor.

Advantage of Permanent Magnet Brushless DC Motor
BLDC motors have many advantages over brushed DC motors and induction motors. A few of these are: • • • • • • Better speed versus torque characteristics Faster dynamic response High efficiency Long operating life Noiseless operation Higher speed ranges

In addition, the ratio of torque delivered to the size of the motor is higher, making it useful in applications where space and weight are critical factors.

Version 2 EE IIT, Kharagpur 3

Structure of Permanent Magnet Brushless DC Motor
BLDC motors come in single-phase, 2-phase and 3-phase configurations. Corresponding to its type, the stator has the same number of windings. Out of these, 3-phase motors are the most popular and widely used. Here we focus on 3-phase motors.

Stator
The stator of a BLDC motor consists of stacked steel laminations with windings placed in the slots that are axially cut along the inner periphery (as shown in Figure 2). Traditionally, the stator resembles that of an induction motor; however, the windings are distributed in a different manner. Most BLDC motors have three stator windings connected in star fashion. Each of these windings are constructed with numerous interconnected coils, with one or more coils are placed in the stator slots. Each of these windings are distributed over the stator periphery to form an even numbers of poles. As their names indicate, the trapezoidal motor gives a back trapezoidal EMF as shown in Figure 35.1.
0 Phase A-B Phase B-C Phase C-A 60 120 180 240 300 360 60

Fig. 35.1 Voltage phase diagram of a 3-phase BLDC motor. In addition to the back EMF, the phase current also has trapezoidal and sinusoidal variations in the respective types of motor. This makes the torque output by a sinusoidal motor smoother than that of a trapezoidal motor. However, this comes with an extra cost, as the sinusoidal motors take extra winding interconnections because of the coils distribution on the stator periphery, thereby increasing the copper intake by the stator windings. Depending upon the power supply capability, the motor with the correct voltage rating of the stator can be chosen. Forty-eight volts, or less voltage rated motors are used in automotive, robotics, small arm movements and so on. Motors with 100 volts, or higher ratings, are used in appliances, automation and in industrial applications.

Version 2 EE IIT, Kharagpur 4

Fig. 35.2 Cross-sectional View of the BLDC motor stator.

Rotor
The rotor is made of permanent magnet and can vary from two to eight pole pairs with alternate North (N) and South (S) poles. Based on the required magnetic field density in the rotor, the proper magnetic material is chosen to make the rotor. Ferrite magnets were traditionally used to make the permanent magnet pole pieces. For new design rare earth alloy magnets are almost universal. The ferrite magnets are less expensive but they have the disadvantage of low flux density for a given volume. In contrast, the alloy material has high magnetic density per volume and enables using a smaller rotor and stator for the same torque. Accordingly, these alloy magnets improve the size-to-weight ratio and give higher torque for the same size motor using ferrite magnets. Neodymium (Nd), Samarium Cobalt (SmCo) and the alloy of Neodymium, Ferrite and Boron (NdFeB) are some examples of rare earth alloy magnets. Figure 35.3 shows cross sections of different arrangements of magnets in a rotor.
N S N S N N S S N S
Circular core with magnets on the periphery Circular core with rectangular magnets embedded in the rotor Circular core with rectangular magnets inserted into the rotor core

N S N

S

S

N

Fig. 35.3 Cross-sections of different rotor cores.

Version 2 EE IIT, Kharagpur 5

Hall Sensors
Unlike a brushed DC motor, the commutation of a BLDC motor is controlled electronically. To rotate the BLDC motor, the stator windings should be energized in a sequence. It is important to know the rotor position in order to understand which winding will be energized following the energizing sequence. Rotor position is sensed using Hall effect sensors embedded into the stator. Most BLDC motors have three Hall sensors embedded into the stator on the non-driving end of the motor. Whenever the rotor magnetic poles pass near the Hall sensors, they give a high or low signal, indicating the N or S pole is passing near the sensors. Based on the combination of these three Hall sensor signals, the exact sequence of commutation can be determined.

Principle of operation and dynamic model of a BLDC Motor
The coupled circuit equations of the stator windings in terms of motor electrical constants are

v a n = R a ia +

v b n = R b ib
v
cn

= R c ic

Ra =Rb =Rc =R

d ( L a a i a + L b a ib + L c a ic ) + e a dt d + ( L a b ia + L b b ib + L c b ic ) + e b dt d + ( L a c ia + L b c ib + L c c ic ) + e c d t

Laa =Lbb =Lcc =Ls Lba =Lab =Lca =Lac =Lbc =Lcb =M
⎡ v an ⎤ ⎡ R ⎢ ⎥ ⎢ ⎢ vbn ⎥ = ⎢ 0 ⎢ v cn ⎥ ⎢ 0 ⎣ ⎦ ⎣ Since, i a + i b 0 R 0 M ⎡ ia ⎤ ⎡ L s ⎢i ⎥ + ⎢ M Ls ⎢ b⎥ ⎢ ⎢ ic ⎥ ⎢ M M ⎣ ⎦ ⎣ + i c = 0 , and with ( L s 0⎤ 0⎥ ⎥ R⎥ ⎦ ⎤ ⎡ ia ⎤ ⎡ e a ⎤ ⎥ d ⎢i ⎥ + ⎢e ⎥ ⎥ dt ⎢ b ⎥ ⎢ b ⎥ ⎥ ⎢ ic ⎥ ⎢ e c ⎥ ⎦ ⎣ ⎦ ⎣ ⎦ M ) = L , we have, M M Ls

⎡ v an ⎤ ⎡ R ⎢v ⎥ = ⎢ 0 ⎢ bn ⎥ ⎢ ⎢ v cn ⎥ ⎢ 0 ⎣ ⎦ ⎣

0 R 0

0 ⎤ ⎡ ia ⎤ ⎡ L 0 ⎥ ⎢ ib ⎥ + ⎢ 0 ⎥⎢ ⎥ ⎢ R ⎥ ⎢ ic ⎥ ⎢ 0 ⎦⎣ ⎦ ⎣

0 L 0

0⎤ ⎡ ia ⎤ ⎡ e a ⎤ ⎥ d ⎢i ⎥ + ⎢e ⎥ 0⎥ b b dt ⎢ ⎥ ⎢ ⎥ ⎢ ic ⎥ ⎢ e c ⎥ L⎥ ⎦ ⎣ ⎦ ⎣ ⎦

where,

R : Stator resistance per phase, assumed to be equal for all phases Ls : Stator inductance per phase, assumed to be equal for all phases. M : Mutual inductance between the phases. ia , ib , ic : Stator current/phase. The instantaneous induced emfs can be written as given in equation (35.1), (35.2) and (35.3). ea = f a (θ r )λ p ωm (35.1)

eb = fb (θr )λ p ωm

(35.2) Version 2 EE IIT, Kharagpur 6

ec = f c (θr )λ p ωm
where, ω
m

(35.3)

is the rotor mechanical speed and θ is the rotor electrical position. r

The machine is represented in the figure 35.4 by a three-phase equivalent circuit, where each phase consists of stator resistance Rs, equivalent self inductance Ls, and a trapezoidal CEMF wave in series. Figure 5 shows the phase diagram of Van, Vbn, Vcn. Id Q1 D1 Q3 b a Q4 G Fig. 35.4 A 3-phase Equivalent Circuit of BLDC motor. +ia D4 Q6 D6 c Q2 D2 c D3 Q5 D5 BLDC Motor b

+ Vd 1GB Switch

-

-

eb e n a Ls + -- + a +
ec

Version 2 EE IIT, Kharagpur 7

Vc Van 0 α Q1 Id P

2π 3

-

2π 3 Q4

Q1

Vbn 0 Q6

Q3

R Q6

Vcn 0

Q5

π 3

Q2

Q2

Fig. 35.5 An Illustration of 3-phase switching sequence. The spatial orientations of the stator MMF vector, under different switching phases of the inverter are shown below in figure 6. Therefore under a cyclic switching scheme one has a rotating stator MMF vector. If the switching can be synchronized with the rotor position, then an approximately fixed angle between the stator flux and the rotor flux can be maintained, while both rotate around the rotor axis. This is similar to the case of the DC motor, where the commutator brush arrangement maintains a fixed spatial direction of the armature flux aligned with the field flux, which is also fixed in space by construction. This is precisely what the inverter switching sequence shown in Fig.5 achieves. The switching instants of the individual transistor switches, Q1 – Q6 with respect to the trapezoidal emf wave is shown in the figure. Note that the emf wave is synchronized with the rotor. So switching the stator phases synchronously with the emf wave make the stator and rotor mmfs rotate in synchronism. Thus, the inverter acts like an electronic commutator that receives switching logical pulses from the rotor position sensor. This is why a BLDC drive is also commonly known as an electronically commutated motor (ECM).

Version 2 EE IIT, Kharagpur 8

T1T2

ca+

b-

T1T4

T3T2 c
-

b+

c+ a-

T5T4 b-

aT3T6 T6T5

Fig. 35.6 Phasor diagram of stator MMF vector. The torque equation is given as,

Te = λ p [ f a (θ r )ia +f b (θ r )ib +f c (θ r )ic ] The equation of motion for simple system is,

Te = J

d ωm +Tl + Bωm dt

where, J is the inertia of the motor and B is the friction coefficient. d ωm 1 => = (T − T − Bωm ) (35.4) dt J e l The relation between angular velocity and angular position (electrical) is given by dθ r P (35.5) = ω dt 2 m

where, P is the number of rotor poles. The state variable ( θr ), rotor position, is required to have the function f a (θr ) , which is given as the trapezoidal function :

f a (θr )= 1
=(
π 6 -θr )* 2 π = -1 π 6 )* 2 π

0 < θr < π 3

π 3 < θr < 2π 3
2π 3 < θr < π

= -1 = (θr -3* =1 Similarly
fb (θr )= f a (θr +2 π ) 3

π < θr < 4π 3
4π 3 < θr < 5π 3
5π 3 < θ r < 2π

Version 2 EE IIT, Kharagpur 9

π ) 3 The induced emfs do not have sharp corners, as is shown in trapezoidal functions, but rounded edges. The emfs are the result of the flux linkages derivatives, and the flux linkages are continuous functions. Fringing also makes the flux density functions smooth with no abrupt edges. It is significant to observe that the phase-voltage equation is identical to the armaturevoltage equation of a dc machine. f c (θr )= f a (θr -2

Points to Ponder: 1
A. Why is a BLDC motor called a dc motor? B. Why does a BLDC motor come with an integrated hall sensor? C. Can you name other purposes which Hall sensors serve in BLDC drives?

Closed Loop Control of PM BLDC Drive
Figure 35.7 represents a position control scheme for PM BLDC motor. For this control scheme speed and position sensors are assumed to be Tachogenerator and LVDT (Linear Variable Differential Transformer). As seen in the figure, there are three loops; the outermost loop called the position loop, the second loop is the speed loop and the innermost loop called the current control loop. The position of the rotor is compared with the reference value, and the rotor position error is amplified through a PD controller. The output of the PD controller is then used as the speed reference command. Based on the speed reference and the speed feedback from a tachogenerator, a PI speed controller generates the torque references, from which, in turn, phase current references are generated.
Current Control Block Inverter PD Controller + BLDC

θref

-

Kp

+
-

Reference Generator
* * * ia , ib , ic

PWMA PWM Regulator PWMB PWMC

Kd

ia, ib, ic

ωm θr

Sensors
Tacho LVDT

Speed

Angle

Fig. 35.7 Position Controlled PMBDCM drive scheme

Version 2 EE IIT, Kharagpur 10

Reference generator

∗ ias

ω

* r

+ -

PI Controller

T e*

1 2λ p

i* p

Reference generator

* i bs

ωr

Reference generator

* i cs

d + filter dt

θr

Fig. 35.8 Block Diagram Representation of speed controlled BLDC drive.

The phase current magnitude ip* command is positive for motoring but negative for regeneration. This signal is then enabled with appropriate polarity to the respective phases with the help of decoder output. The actual phase currents track the command currents by hysteresis-band current control. At any instant, two phase currents are enabled, one with positive polarity and another with negative polarity. Consider, for example, the motoring mode time duration when phase a positive current +ia* and phase b negative current –ib* commands are enabled by the decoder. Devices Q1 in phase a and Q6 in phase b are turned on simultaneously to increase +ia and –ib respectively. When the currents (equal in magnitude) tend to exceed the hysteresis band, Q6 is turned off. The decaying freewheeling current will flow then through D3 and Q1.In this case, three upper devices (Q1, Q3, and Q5) of the inverter are turned on sequentially in the middle of the respective positive voltage half-cycles, whereas the lower devices (Q4, Q6, and Q2) are chopped in sequence for 2π/3 angles in respective negative voltage half-cycles with the help of a decoder for controlling the current ip*.

Points to Ponder: 2
A. Why is a PD controller used in the position loop ? B. Explain how the current control loop works

Typical BLDC Motor Applications
BLDC motors find applications in every segment of the market. Automotive, appliance, industrial controls, automation, aviation and so on, have applications for BLDC motors. Out of these, we can categorize the type of BLDC motor control into three major types: • Constant load • Varying loads • Positioning applications Version 2 EE IIT, Kharagpur 11

Applications with Constant Loads
These are the types of applications where variable speed is more important than keeping the accuracy of the speed at a set speed. In addition, the acceleration and deceleration rates are not dynamically changing. In these types of applications, the load is directly coupled to the motor shaft. For example, fans, pumps and blowers come under these types of applications. These applications demand low-cost controllers, mostly operating in open-loop.

Applications with Varying Loads
These are the types of applications where the load on the motor varies over a speed range. These applications may demand a high-speed control accuracy and good dynamic responses. In home appliances, washers, dryers and compressors are good examples. In automotive, fuel pump control, electronic steering control, engine control and electric vehicle control are good examples of these. In aerospace, there are a number of applications, like centrifuges, pumps, robotic arm controls, gyroscope controls and so on. These applications may use speed feedback devices and may run in semi-closed loop or in total closed loop. These applications use advanced control algorithms, thus complicating the controller. Also, this increases the price of the complete system.

Positioning Applications
Most of the industrial and automation types of application come under this category. The applications in this category have some kind of power transmission, which could be mechanical gears or timer belts, or a simple belt driven system. In these applications, the dynamic response of speed and torque are important. Also, these applications may have frequent reversal of rotation direction. A typical cycle will have an accelerating phase, a constant speed phase and a deceleration and positioning phase. The load on the motor may vary during all of these phases, causing the controller to be complex. These systems mostly operate in closed loop. There could be three control loops functioning simultaneously: Torque Control Loop, Speed Control Loop and Position Control Loop. Optical encoder or synchronous resolvers are used for measuring the actual speed of the motor. In some cases, the same sensors are used to get relative position information. Otherwise, separate position sensors may be used to get absolute positions. Computer Numeric Controlled (CNC) machines are a good example of this. Process controls, machinery controls and conveyer controls have plenty of applications in this category.

Points to Ponder: 3
A. What are the reasons behind the popularity of BLDC motors over dc motors for servo applications? B. Name application areas where a BLDC motor drive is preferable over an ac drive.

Lesson Summary
In this lesson, the following topics related to PM BLDC motor have been discussed. A. Advantages of PM BLDC Motor. Version 2 EE IIT, Kharagpur 12

B. Structure of PM BLDC Motor. C. Principle of operation and dynamic model of a BLDC Motor. D. Closed Loop Control of PM BLDC Drive. E. Applications of Typical BLDC Motor.

Version 2 EE IIT, Kharagpur 13

Answers, Remarks and Hints to Points to Ponder Points to Ponder: 1
A. Why is a BLDC motor called a dc motor?
Ans: Because in a BLDC motor the stator and the rotor fluxed maintain a constant angle, although both are rotating in space. This is similar to the dc motor, but in the dc motor, both the armature and the field flux are stationary in space.

B. Why does a BLDC motor come with an integrated hall sensor?
Ans: Because, to maintain a constant angle difference between the stator and the rotor flux, the stator current has to be switched from winding to winding depending on the rotor position. So the rotor position is sensed in all BLDC motors, by the Hall sensors.

C. Can you name other purposes which Hall sensors serve in BLDC drives?
Ans: Hall sensors are also used in many drives to sense stator current to implement a current loop for fast torque control.

Points to Ponder: 2
A. Why is a PD controller used in the position loop ?
Ans: Since the velocity to position dynamics already contains an integral dynamics, no integrator needs to be introduced n the controller for achieving zero steady state error. Thus, to improve the control bandwidth, only a PD controller is used.

B. Explain how the current control loop works.
Ans: As explained for dc motors in an earlier lesson, switching the supply across the BLDC motor makes the current increase linearly, since the back emf, being proportional to speed can be assumed to remain constant. When the current crosses upper trip point of a hysteresis block set around the current set point, the supply is switched off, making the current free wheel through one transistor and one diode, and reduce linearly. Again, when the current crosses the lower trip point, the supply is switched on. Thus, the current oscillates at the PWM switching frequency around the current set point.

Points to Ponder: 3
A. What are the reasons behind the popularity of BLDC motors over dc motors for servo applications?
Ans: Higher torque to weight ratio, due to rare earth magnets which have less weight and have much higher operating flux density limits. This leads to faster dynamic response. Also, the lack of brushes and commutators reduce maintenance problems.

Version 2 EE IIT, Kharagpur 14

B. Name application areas where a BLDC motor drive is preferable over an ac drive.
Ans: Most control applications are examples, such as, CNC control, Avionic controls etc.

Version 2 EE IIT, Kharagpur 15

Module 8
Industrial Embedded and Communication Systems
Version 2 EE IIT, Kharagpur 1

Lesson 36
Introduction to Real Time Embedded Systems
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
After going through this lesson the student would be able to • • • Define a Real Time Embedded System Describe major hardware components of an Embedded system. Describe typical architectures for such systems

Introduction Definition
A Real Time Embedded System (RTES) an RTES is essentially a special-purpose computational system built into another machine with the sole purpose of controlling the embedder. Below we explain the key terms used in the above definition.

Real Time
“Real’-time usually means time as prescribed by external sources”. For example the time struck by clock (however fast or late it might be). The timings are generated by the requirements of the over all system. This external timing requirements imposed by the user is the real-time for the embedded system.

Embedded
Embedded, as per the Collins Cobuild dictionary, means, “firmly and deeply fixed within a surrounding mass”. Thus, an embedded system is deep within a surrounding system so that, from an external appearance such a system is not visible. For example, a CNC system controller is not apparent in the overall machine. On the other hand, a standalone system such as the PC does not have any such surrounding and is visible by itself. The system is firmly fixed, in the sense that it is an integral part of the overall system, both functionally and structurally. Thus “A Real Time Embedded System” (RTES), in our context, is a computational subsystem within an overall system built to discharge a specific industrial task (such as precision machining), that is to be carried out in a definite manner with respect to time. Henceforth we refer to them as RTES.

Version 2 EE IIT, Kharagpur 3

Technical Features

• Shaft encoder input • Onboard dedicated high speed CPU • Motor drive set point output • Digital IO • Programmer port • Set point from data bus

Fig. 36.1 Embedded Single Board Position Controller Module RTES as a generic term may mean a wide variety of systems in the real world. However we will be concerned about those, which use programmable devices such as microprocessors or microcontrollers and have specific functions. We shall also restrict ourselves mainly to industrial RTES applications although not explicitly so. Much of the discussion is also applicable to other application domains of RTES, and is therefore presented in a general setting.

Typical Characteristics of an RTES Single-Functioned
Here “single-functioned” means specific functions. The RTES is usually meant for very specific functions. Generally a special purpose microprocessor executes a program over and over again for a specific purpose. The same program is executed repeatedly throughout almost the whole life of the systems, unless reprogrammed for upgrading. These operations are monitored and controlled by an operating system called as Real Time Operating System (RTOS) which is simpler, than a general purpose OS like Windows because these are built for a specific set of functionality, use a specific set of computing resources and try to achieve specific performance metrics.

Tightly Constrained
The constraints on the design of RTES are more severe than their non-real-time non-embedded counter parts. Time-domain constraints are the first thing that is taken care while developing such a system. Size, weight, power consumption and cost are the other major factors. However, RTES often turned out to be quite complex to achieve performance in the face of resource and timing constraints. The major burden of complexity often goes to the software.

Version 2 EE IIT, Kharagpur 4

Reactive and Real Time
Many embedded systems must continually react to events and processes in the system’s environment and must compute results in real time. For example, an industrial controller continually monitors and reacts to process the signals generated by analog and digital sensors. It must compute control inputs and must generate outputs repeatedly within a limited time to actuators and indicators; a delayed computation could result in a failure to maintain control of the industrial process. In contrast, a desktop computer system typically focuses on computations, with relatively infrequent (from the computer’s perspective) reactions to input devices. In addition, a delay in those computations, while perhaps inconvenient to the user, typically does not result in a system failure.

Point to Ponder: 1
A. Give one example of an embedded system and of one that is not. Justify your examples. B. Name an industrial embedded system. Explain whether the above characteristics hold for it. C. Name an Industrial Real Time systems, which may not be called embedded. Explain your answer.

Common Architecture
Unlike general-purpose computers, a generic architecture cannot be rigidly defined for Real Time Embedded System. There are many types of architecture proposed by manufacturers of a spectacular variety of embedded processors. However for the sake of our understanding we can discuss some common form of systems at the block diagram level. Any system can hierarchically divided into subsystems. Each sub-system may be further segregated into smaller systems. Each of these smaller systems, in turn, consists of discrete parts. Together, this is referred to as the hardware configuration. Some of these parts may be programmable and therefore must have memory to store the programs. In an RTES the on-chip or on-board non-volatile memory is used to store these programs. These programs constitute both a Real Time Operating System (RTOS) as well as the application programs. The overall software system continually runs, as long as the RTES receives power. Particularly, for industrial applications, embedded systems often runs for days and even months on end.

Version 2 EE IIT, Kharagpur 5

EMBEDDED SYSTEM hardware software mechanical optical … subsystem digital subsystem sensors actuators

analog subsystem

Fig. 36.2 Typical Embedded System Architecture

Components of an Embedded System
Architecture of a typical embedded-system is shown in Fig. 36.2. The hardware unit consists of the above units along with a digital as well as an analog subsystem. The software in the form of a RTOS resides in the memory. A typical embedded system consists, by and large, of the following units housed on a single board or chip. 1. Processor 2. Memory 3. Input/Output devices and interface chips 4. Real Time System Operating System Software 5. Application Software

1. Microprocessor
The microprocessors used in RTES are different from the general-purpose microprocessors like Pentium etc. They are often designed to meet specific application requirements. For example, a special function module for a PLC card may contain a special processor for high speed, real time, multi-channel I/O.

2. Memory
The microprocessor and memory must co-exit on the same PCB or same chip. Compactness, speed and low power consumption are the characteristics required for the memory to be used in an RTES. For housing the operating system, Read Only Memory (ROM) is used. Sometimes it is required to load-reload programs more often than never. The program or data loaded might exist for considerable duration. In these cases the memory should be capable of retaining the information even after the power is removed. In other words the memory should Version 2 EE IIT, Kharagpur 6

be non-volatile and should be easily programmable too. It is achieved by using Flash memories.

3. Input Output Devices and Interfaces
Input/Output interfaces are necessary to make the RTES interact with the external world. This is one of the most important subsystems for industrial RTES. The RTES needs to interact with a wide variety of sensors and actuators. These RTES should also have standard hardware interfaces to other devices such as Desktop Computers, Local Area Networks (LAN) and other RTES. These input/output devices along with standard software protocols in the RTOS provide the necessary interface to these standards.

4. Software
The software in an RTES can be classified as System and Application Software. A Real Time Operating System sitting on the non-volatile memory of the RTES, The application software are essentially cyclic or interrupt driven programs that execute read-compute-write cycles. Multiple such programs execute under a priority driven scheduling policy managed by the RTOS. These issues are discussed in the next lesson. Besides the above, an RTES may have various other components and Application Specific Integrated Circuits (ASIC) for specialized functions such as motor control or counting. In the sequel, in this lesson, we shall discuss the main hardware components of an RTES.

Point to Ponder: 2
A. State one characteristics each of industrial embedded systems in respect of Processors, Memory and I/O. B. Describe typical i/o organization of industrial embedded systems.

Processors
The central processing unit is the most important component in an embedded system. It exists in an integrated manner along with memory and other peripherals. Depending on the type of applications the processors are broadly classified into 3 major categories 1. General Purpose Microprocessors 2. Microcontrollers 3. Digital Signal Processors In most of the applications, the design is carried out using already available processors in the market. However, the Field Programmable Gate Arrays (FPGA) can be used to implement simple customized processors easily. An FPGA is a type of logic chip that can be programmed as a sequential machine. FPGA’s support thousands of gates which can be connected and disconnected like an EPROM (Erasable Programmable Read Only Memory). They are especially

Version 2 EE IIT, Kharagpur 7

popular for prototyping integrated circuit designs. Once the design is set, hardwired chips are produced for faster performance.

General Purpose Processors
A general-purpose processor is designed to solve problems in a large variety of applications as diverse as communications, automotive and industrial embedded systems. These processors are generally inexpensive because these are manufactured in large quantities. The NRE (Nonrecurring Engineering Cost) is, therefore, spread over a large number of units. Being cheaper the manufacturer can invest more for improving the VLSI design with advanced optimized architectural features. Thus the performance, size and power consumption can be improved. Also the supporting hardware is cheap and easily available. However, only a part of the processor capability may be needed for a specific design and hence the over all embedded system will not be as optimized as it should have been as far as the space, power and reliability is concerned. Pentium IV is such a general purpose processor with most advanced architectural features. Compared to its overall performance the cost is also low. A general purpose processor consists of a data path and a control unit tightly linked with the memory. The Data Path consists of circuitry for transforming data and storing temporary data. It contains an arithmetic-logic-unit (ALU) capable of transforming data through operations such as addition, subtraction, logical AND, logical OR, inverting, shifting etc. The data-path also connects registers capable of storing temporary data generated out of ALU or related operations. The internal data-bus carries data within the data path while the external data bus carries data to and from the data memory. The size of the data path indicates the bit-size of the CPU (e.g. 8085 has an 8-bit data path). The Control Unit consists of circuitry for retrieving program instructions and for moving data to, from, and through the data-path according to those instructions. It has a program counter (PC) to hold the address of the next program instruction to fetch and an Instruction register(IR) to hold the fetched instruction. It also has a timing unit in the form of state registers and control logic. The controller sequences through the states and generates the control signals necessary to read instructions into the IR and control the flow of data in the data path. Generally the size of the address space is dictated by the capacity of the control unit as it is responsible to communicate with the memory. For each instruction the controller typically sequences through several stages, such as fetching the instruction from memory, decoding it, fetching the operands, executing the instruction in the data path and storing the results. Each stage takes a number of clock cycles.

Version 2 EE IIT, Kharagpur 8

Processor
Control Contr ol/Stat

Datapath AL

Controller

Registers P IR

I/ Memory Fig. 36.3 The architecture of a General Purpose Processor

Microcontroller
Just as you put all the major components of a Desktop PC on to a Single Board Computer (SBC) if you put all the major components of a Single Board Computer on to a single chip it will be called as a Microcontroller. Because of the limitations in the VLSI design most of the input/output functions exist in a simplified manner. Typical architecture of such a microprocessor is shown in Fig.36.4.

Serial Port Parallel Port

IRAM

Address Bus

Interrupt Controller

XRAM

Peripheral Bus

Timers Parallel Port

C500 Core (1 or 8 Datapointer)

ROM

Control

Access Control

A D

Housekeeper

MDU Port0/Port2 WDU

Fig. 36.4 The architecture of a typical microcontroller named as C500 from Infineon Technology, Germany Version 2 EE IIT, Kharagpur 9

Data Bus

Ext. Control

RST EA PSEN ALE XTAL

The various units of the processors are as follows: The C500 Core contains the CPU which consists of the Instruction Decoder, ALU and Program Control section The housekeeper unit generates internal signals for controlling the functions of the individual internal units within the microcontroller. Port 0 and Port 2 are required for accessing external code and data memory and for emulation purposes. The external control block handles the external control signals and the clock generation. The access control unit is responsible for the selection of the on-chip memory resources. The IRAM provides the internal RAM, which includes the general purpose, registers. The XRAM is another additional internal RAM provided sometimes Interrupt requests from peripheral units are handled by an Interrupt Controller Unit. Serial interfaces, timers, capture/compare units, A/D converters, watchdog units (WDU), or a multiply/divide unit (MDU) are typical examples for on-chip peripheral units. The external signals of these peripheral units are available at multifunctional parallel I/O ports or at dedicated pins.

Digital Signal Processor (DSP)
These processors have been designed based on the modified Harvard Architecture to handle real time signals. The features of these processors are suitable for implementing signal processing algorithms. One of the common operations required in such applications is array multiplication. For example convolution and correlation require array multiplication. This is accomplished by multiplication followed by accumulation and addition. This is generally carried out by Multiplier and Accumulator (MAC) units. Sometimes it is known as MACD, where D stands for Data move. Generally all the instructions are executed in single cycle.

Processing Unit

Result/Operands

Data Memory

Status

Opcode

Address

Control Unit

Instructions Address

Program Memory

Fig. 36.5 The modified Harvard architecture Version 2 EE IIT, Kharagpur 10

The MACD type of instructions can be executed faster by parallel implementation. This is possible by separately accessing the program and data memory in parallel. This can be accomplished by the modified architecture shown in Fig.3. These DSP units generally use Multiple Access and Multi Ported Memory units. Multiple access memory allows more than one access in one clock period. The Multi-ported Memory allows multiple addresses as well Data ports. This also increases the number of access per unit clock cycle. Address Bus 1 Data Bus 1

Address Bus 2

Dual Port Memory

Data Bus 2

Fig. 36.6 Dual Ported Memory The Very Long Instruction Word (VLIW) architecture is also suitable for Signal Processing applications. This has a number of functional units and data paths as seen in Fig.5. The long instruction words are fetched from the memory. The operands and the operation to be performed by the various units are specified in the instruction itself. The multiple functional units share a common multi-ported register file for fetching the operands and storing the results. Parallel random access to the register file is possible through the read/write cross bar. Execution in the functional units is carried out concurrently with the load/store operation of data between RAM and the register file

Multi-ported Register File

Program Control Unit

Read/Write Cross Bar

Functional Unit 1

.......

Functional Unit n

Instruction Cache
Fig. 36.7 Block Diagram of VLIW architecture

Version 2 EE IIT, Kharagpur 11

Microprocessors vs Microcontrollers
A microprocessor is a general-purpose digital computer’s central processing unit. To make a complete microcomputer, memory (ROM and RAM) memory decoders, an oscillator, and a number of I/O devices are added. The prime use of a microprocessor is to read data, perform extensive calculations on that data, and store the results in a mass storage device or display the results. These processors have complex architectures with multiple stages of pipelining and parallel processing. The memory is divided into stages such as multi-level cache and RAM. The development time of General Purpose Microprocessors is high because of a very complex VLSI design. The design of the microcontroller is driven by the desire to make it as expandable and flexible as possible. Microcontrollers usually have on chip RAM and ROM (or EPROM) in addition to on chip i/o hardware to minimize chip count in single chip solutions. As a result of using on chip hardware for I/O and RAM and ROM they usually have pretty low performance CPU. Microcontrollers also often have timers that generate interrupts and can thus be used with the CPU and on chip A/D D/A or parallel ports to get regularly timed I/O. The prime use of a microcontroller is to control the operations of a machine using a fixed program that is stored in ROM and does not change over the lifetime of the system. The microcontroller is concerned with getting data from and to its own pins; the architecture and instruction set are optimized to handle data in bit and byte size. The contrast between a microcontroller and a microprocessor is best exemplified by the fact that most microprocessors have many operation codes (opcodes) for moving data from external memory to the CPU; microcontrollers may have one or two. Microprocessors may have one or two types of bit-handling instructions; microcontrollers will have many.

Microprocessors vs. DSP
DSPs are microprocessors specialized for signal processing applications. They are designed following a Harvard architecture, contrasted with the Von Neuman architecture, shown below. Program Memory Processor Memory Processor Data Memory

Fig. 36.8 The Von Neumann and the Harvard architechtures DSPs, make concurrent, multiple memory accesses per cycle. GPPs use a multi-level hierarchy of memory, with an elaborate virtual memory system. Dedicated DSP hardware performs all key arithmetic operations in 1 cycle. This involves features such as, complex addressing of multiple data streams for each instruction, hardware loop control unit, typically unavailable for GPPs. DSPs support a narrower range of on-chip peripherals.

Point to Ponder: 3

Version 2 EE IIT, Kharagpur 12

A. Give one example each of an embedded system that can use a GPP, a microcontroller and a DSP. Justify your examples. B. Name one each of commercial versions of a GPP, a microcontroller and a DSP that is used in industrial embedded applications.

Memory
Memory serves processor’s short and long-term information storage requirements while registers serve the processor’s short-term storage requirements. Both the program and the data are stored in the memory. In the Princeton Architecture the data and program occupy the same physical memory. In Harvard Architecture the program and the data occupy separate memory blocks. The former leads to simpler architecture. The later needs two separate connections and hence the data and program can be made parallel leading to parallel processing. General-purpose processors typically have a Princeton Architecture. The memory may be Read-Only-Memory or Random Access Memory (RAM). It may exist on the same chip with the processor itself or may exist outside the chip. The on-chip memory is faster than the off-chip memory. To reduce the access (read-write) time a local copy of a portion of memory can be kept in a small but fast memory called the cache memory. The memory also can be categorized as Dynamic or Static. Dynamic memory dissipate less power and hence can be compact and cheaper. But the access time of these memories are slower than their Static counter parts. In Dynamic RAMs (or DRAM) the data is retained by periodic refreshing operation. While in the Static Memory (SRAM) the data is retained continuously. SRAMs are much faster than DRAMs but consume more power. The intermediate cache memory is an SRAM. When the CPU needs data, it first looks in its own data registers. If the data isn’t there, the CPU looks to see if it’s in the nearby Level 1 cache. If that fails, it’s off to the Level 2 cache. If it’s nowhere in cache, the CPU looks in main memory. If not there, the CPU gets it from disk. All the while, the clock is ticking, and the CPU is sitting there waiting.

Point to Ponder: 4
A. State with justification if the following statements are right (or wrong) a. Cache memory can be a static RAM b. Dynamic RAMs occupy more space per word storage c. The full-form of SDRAM is static-dynamic RAM d. BIOS in your PC is not a Random Access Memory (RAM) B. Order the following in the increasing order of their access speed Flash Memory, Dynamic Memory, Cache Memory, CDROM, Hard Disk, Magnetic Tape, Processor Memory

Input/Output Devices and Interface Chips
Typical RTES interact with the environment and users through some inbuilt hardware. Occasionally external circuits are required for communicating with user, other computers or a network. To generate an analog signal from the microprocessor we need a Digital to Analog Converter (DAC) and to accept analog signal we need and Analog to Digital Converter (ADC). These DAC Version 2 EE IIT, Kharagpur 13

and ADC again have certain control modes. They may also operate at different speed than the microprocessor. To synchronize and control these interface chips we may need another interface chip. Similarly we may have interface chips for keyboard, screen and antenna. These chips serve as relaying units to transfer data between the processor and input/output devices. The input/output devices are generally slower than the processor. Therefore, the processor may have to wait till they respond to any request for data transfer. Number of idle clock cycles may be wasted for doing so. However, the input-output interface chips carry out this task without making the processor to wait or idle. Signal Conditioning and Amplification Amplification A-D Converter Processor Actuator D-A Converter Memory

Sensor

Fig. 36.9 The typical input/output interface blocks Besides the above units some real time embedded systems may have specific circuits included on the same chip or circuit board. They are known as Application Specific Integrated Circuit (ASIC). Some examples are 1. Filters: Filters are used to condition the incoming signal by eliminating the out-band noise and other unnecessary signals. A specific class of filters called Anti-aliasing filters, are used before the A-D converters to prevent aliasing while acquiring a broad-band signal. 2. Controllers: These are specific circuits for controlling, motors, actuators and light-intensities etc.

Lesson Summary
The following topics were covered. a. Definition and characteristics of RTES b. Architecture and Components of RTES c. Processors d. Memory e. I/O and interface modules

Version 2 EE IIT, Kharagpur 14

Answers, Remarks and Hints to Points to Ponder Point to Ponder: 1
A. Give one example of an embedded system and of one that is not. Justify your examples. Ans: The controller for a microwave oven is an embedded system. Its existence is deep within the oven casing and not immediately visible at all. However, it controls each functionality of the microwave oven. On the other hand a pocket calculator is not an embedded system although it has very similar structural characteristics and handles similar kind of i/o, like LCD display and keypad. The calculator is a standalone computing device, which does not have a embedding environment. B. Name an industrial embedded system. Explain whether the above characteristics hold for it. Ans: Consider a CNC milling machine controller. One such machine is shown. Note that the controller is not seen on the machine, although a display and a keyboard

are. The controller is custom designed to control the CNC milling machine. It cannot be used, as a PC for example. Thus it is single-functioned. It is clearly reactive and real-time, since it interacts with sensors every sampling cycle. Its design is likely to be constrained by cost. However, design of industrial systems is less constrained compared to other categories, in respect of power, size, weight and consequently, memory, processor speed etc.

Point to Ponder: 2
A. State one characteristics each of industrial embedded systems in respect of Processors, Memory and I/O. Ans: Processors for industrial embedded systems are generally characterized by a sophisticated interrupt handling system. This is necessary to support an asynchronous i/o subsystem, that is typical of industrial i/o as well as to facilitate a preemptive priority scheduler that is an integral part of an RTOS. Memory configurations for industrial Version 2 EE IIT, Kharagpur 15

embedded systems are characterized by the fact that there is generally enough available memory to preclude a virtual memory system. This is because virtual memory systems can introduce significant variations in task latencies and are therefore not good reliable for realtime applications. One important characteristics of industrial embedded i/o is that it involves both processor driven cyclic i/o as well device interrupt driven asynchronous i/o. B. Describe typical i/o organization of industrial embedded systems. Ans: I/O in industrial embedded systems can be divided into field i/o and device i/o. Field i/o typically means signals from sensors and signals to actuators and indicators. These are generally sensed cyclically, as for every scan of an RLL. Other kinds of i/o involve data exchange with various system devices such as programmers, MMI devices or communication processors, keyboards and displays. Finally the last kind of i/o involves network i/o following standard protocols such as the Ethernet or the fieldbus.

Point to Ponder: 3
A. Give one example each of an embedded system that can use a GPP, a microcontroller and a DSP. Justify your examples. Ans: A large industrial controller like a rack-based PLC can use a GPP. Motor drive controllers often use DSPs. A single-loop temperature controller can use a microcontroller. These are chosen depending on the number of computing tasks, special kind of floating point computational requirements and the need for a simple system with limited off-chip resource requirement. B. Name one each of commercial versions of a GPP, a microcontroller and a DSP that is used in industrial embedded applications. Ans: 80186 processors have been used in the past to build embedded controller cards. More recently processors like the Power-PC may be used for industrial applications. Similarly, 8031 processors have been used for building small industrial controllers. More recently one can use ARM processors to build industrial embedded devices like intelligent sensors. Texas instruments has a special category of DSPs that have been designed for motor control (TMS 658030)

Point to Ponder: 4
A. State with justification if the following statements are right (or wrong) a. Cache memory can be a static RAM b. Dynamic RAMs occupy more space per word storage c. The full-form of SDRAM is static-dynamic RAM d. BIOS in your PC is not a Random Access Memory (RAM) Ans: a. Yes, because it is very fast and limited in density; b. No, they have the highest densities among memory technologies; c. No. it is synchronous dynamic RAM; d. Yes, BIOS cannot be on volatile memory like RAM. Version 2 EE IIT, Kharagpur 16

B. Order the following in the increasing order of their access speed Flash Memory, Dynamic Memory, Cache Memory, CDROM, Hard Disk, Magnetic Tape, Processor Memory Ans: CDROM, Magnetic tape, Hard disk, Flash memory, Dynamic memory, Cache memory, Processor memory

Version 2 EE IIT, Kharagpur 17

Module 8
Industrial Embedded and Communication Systems
Version 2 EE IIT, Kharagpur 1

Lesson 37
Real-Time Operating Systems: Introduction and Process Management
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
After learning the lesson the student should be able to: A. Describe the major functions of an Operating System B. Define multi-tasking and describe its advantages C. Describe the task states and transitions in the execution life cycle under a multi-tasking OS D. Define the concept of preemptive priority scheduling E. Describe common multi-tasking architectures of RTOS. F. Describe the classification of computing tasks in terms of their timing constraints

Introduction
Embedded Computing Applications exist in a spectacular range of size and complexity for areas such as home automation to cell phones, automobiles and industrial controllers. Most of these applications demand such functionality, performance and reliability from the software that simple and direct assembly language programming of the processors is clearly ruled out. Moreover, as a distinguishing feature from general purpose computing, a large part of the computation is “real-time” or time constrained and also reactive or “external event driven” since such systems generally interface strongly with the external environment through a variety of devices. Thus, an operating system is generally used. An operating system facilitates development of application program by making available a number of services, which, otherwise would have to be coded by the application program. The application programs “interface” with the hardware through a operating system services and functions. It is therefore important understand the basic features of such operating systems. In this chapter we present the fundamental concepts of a real-time operating system in a generic context. Most large industrial controllers employ such operating systems. Before we undertake the discussion, we briefly review the nature of computation in industrial automation systems.

Nature of IA computation
Computation for industrial automation are typified by the following characteristics: • The computation is reactive in nature, that is, the computation is carried out with respect to real-world physical signals and events. Therefore industrial computers must have extensive input-output subsystems that interface it with physical signals In general speed of computation is not very high. This is because dynamics for variables such as temperature are very slow. For electromechanical systems speed requirement is higher. However, there processing is often done by dedicated coprocessors. The algorithmic complexity of computation is generally not very high, although it is increasing due to the deployment of sophisticated control and signal processing algorithms. Therefore general-purpose processors are sufficient, almost in all cases.

Version 2 EE IIT, Kharagpur 3

A large part of the computational tasks are repetitive or cyclic, that is their executions are invoked periodically in time. Many of these tasks are also real-time, in the sense that there are deadlines associated with their execution. When there are a large number of such tasks to be executed on a single processor, appropriate scheduling strategies are required. Since the number of tasks is mostly fixed, simple static scheduling policies are generally adequate. Although computational demands are not very high, the computation is critical in nature. Indeed, the cost ratios of industrial controllers and the plant equipment they control can well be in excess of one thousand. This requirement is coupled with the fact that such controllers are often destined to work in harsher industrial environments. Thus, reliability requirement of industrial control computers are very high. In fact, this is one of the main reasons that decide the high price of such systems compared to commercial computers of comparable or even higher data processing capability.

In this lesson the basic features of a Real-Time Operating System (RTOS) are introduced and motivated from the demands of typical real-time applications. Before the RTOS is discussed the typical functionalities of a general Operating System (OS) reviewed briefly. While the RTOS has some distinguishing features, it is also largely similar to general purpose operating systems. The discussion here is necessarily brief and user oriented, rather than designer oriented. For a detailed exposition on OS any standard textbook on the subject may be referred to.

Operating Systems (OS) Basics
An Operating System is a collection of programs that provides an interface between application programs and the computer system (hardware). Its primary function is to provide application programmers with an abstraction of the system resources, such as memory, input-output and processor, which enhances the convenience, efficiency and correctness of their use. These programs or functions within the OS provide various kinds of services to the application programs. The application programs, in turn, call these programs to avail of such services. Thus the application programs can view the computer resources as abstract entities, (for example a block of memory can be used as a named sequential file with the abstract Open, Close, Read, Write operations) without need for knowing the low level hardware details (such as the addresses of the memory blocks). To get a better idea of such services provided by the OS, the user may like to refer to the DOS services for the IBM PC Compatibles. The natural way to view computation in a typical modern computer system is in the form of a number of different programs, all of which, apparently, run in parallel. However, very often, all the programs in the system are executed on a single physical CPU or processor. Each program runs for a small continuous duration at a time, before it is stopped and another program begins to execute. If this is done rapidly enough, it appears as if all programs are running simultaneously. These programs very often perform independent computations, such as the programs executing in different windows on a PC. In real time systems, most often, such programs cooperate with each other, by exchanging data and synchronizing each other’s execution, to achieve the overall functionality and performance of the system.

Version 2 EE IIT, Kharagpur 4

Types of Operating Systems
• • • Stand-Alone Operating systems Network Operating systems Embedded Operating Systems

Stand-Alone Operating System
It is a complete operating system that works on a desktop or notebook computer. Examples of stand-alone operating systems are: • DOS • Windows 2000 Professional • Mac OS X

Network Operating System
It is an operating system that provides extensive support for computer networks. A network operating system typically resides on a server. Examples of a network operating system are: • Windows 2000 Server • Unix • Linux • Solaris

Embedded Operating System
You can find this operating system on handheld computers and small devices, it resides on a ROM chip. Examples of embedded operating systems are: • Windows CE • Pocket PC 2002 • Palm OS The above classification is based on the computing hardware environment towards which the OS is targetted. All three types can be either of a real-time or a non-real-time type. For example, VxWorks is an RTOS of the first category, RT-Linux and ARTS are of the second category and Windows CE of the third category.

Point to Ponder 1
A. What are the factors on which the execution time of a program depends on? B. While a task is executing, is the CPU continuously busy? C. Suppose there are two mutually independent programs A and B. If only program A (B) executes on the processor P, it takes tA ( tB) units of time. If these programs are fired simultaneously and run under multi-programming, neglecting time related to program switching, would the execution time for both programs taken together be greater than, equal to or less than (tA + tB )?
B B

Version 2 EE IIT, Kharagpur 5

Real-Time Operating Systems
A Real-Time OS (RTOS) is an OS with special features that make it suitable for building realtime computing applications also referred to as Real-Time Systems (RTS). An RTS is a (computing) system where correctness of computing depends not only on the correctness of the logical result of the computation, but also on the result delivery time. An RTS is expected to respond in a timely, predictable way to unpredictable external stimuli. Real-time systems can be categorized as Hard or Soft. For a Hard RTS, the system is taken to have failed if a computing deadline is not met. In a Soft RTS, a limited extent of failure in meeting deadlines results in degraded performance of the system and not catastrophic failure. The correctness and performance of an RTS is therefore not measured in terms of parameters such as, average number of transactions per second as in transactional systems such as databases. A Good RTOS is one that enables bounded (predictable) behavior under all system load scenarios. Note however, that the RTOS, by itself cannot guarantee system correctness, but only is an enabling technology. That is, it provides the application programmer with facilities using which a correct application can be built. Speed, although important for meeting the overall requirements, does not by itself meet the requirements for an RTOS.

Programs, Processes, Tasks and Threads
The above four terms are often found in literature on OS in similar contexts. All of them refer to a unit of computation. A program is a general term for a unit of computation and is typically used in the context of programming. A process refers to a program in execution. A process is an independently executable unit handled by an operating system. Sometimes, to ensure better utilisation of computational resources, a process is further broken up into threads. Threads are sometimes referred to as lightweight processes because many threads can be run in parallel, that is, one at a time, for each process, without incurring significant additional overheads. A task is a generic term, which, refers to an independently schedulable unit of computation, and is used typically in the context of scheduling of computation on the processor. It may refer either to a process or a thread.

Multitasking
A multitasking environment allows applications to be constructed as a set of independent tasks, each with a separate thread of execution and its own set of system resources. The inter-task communication facilities allow these tasks to synchronize and coordinate their activity. Multitasking provides the fundamental mechanism for an application to control and react to multiple, discrete real-world events and is therefore essential for many real-time applications. Multitasking creates the appearance of many threads of execution running concurrently when, in fact, the kernel interleaves their execution on the basis of a scheduling algorithm. This also leads to efficient utilisation of the CPU time and is essential for many embedded applications where processors are limited in computing speed due to cost, power, silicon area and other constraints. In a multi-tasking operating system it is assumed that the various tasks are to cooperate to serve the requirements of the overall system. Co-operation will require that the tasks communicate with each other and share common data in an orderly an disciplined manner, without creating Version 2 EE IIT, Kharagpur 6

undue contention and deadlocks. The way in which tasks communicate and share data is to be regulated such that communication or shared data access error is prevented and data, which is private to a task, is protected. Further, tasks may be dynamically created and terminated by other tasks, as and when needed. To realise such a system, the following major functions are to be carried out. A. Process Management • • interrupt handling task scheduling and dispatch • • create/delete, suspend/resume task manage scheduling information – priority, scheduling policy, etc B. Interprocess Communication and Synchronization • • • Code, data and device sharing Synchronization, coordination and data exchange mechanisms Deadlock and Livelock detection

C. Memory Management • dynamic memory allocation • memory locking • Services for file creation, deletion, reposition and protection D. Input/Output Management • Handles request and release functions and read, write functions for a variety of peripherals

The following are important requirements that an OS must meet to be considered an RTOS in the contemporary sense. A. The operating system must be multithreaded and preemptive. e.g. handle multiple threads and be able to preempt tasks if necessary. B. The OS must support priority of tasks and threads. C. A system of priority inheritance must exist. Priority inheritance is a mechanism to ensure that lower priority tasks cannot obstruct the execution of higher priority tasks. D. The OS must support various types of thread/task synchronization mechanisms. E. For predictable response : a. The time for every system function call to execute should be predictable and independent of the number of objects in the system. b. Non preemptable portions of kernel functions necessary for interprocess synchronization and communication are highly optimized, short and deterministic

Version 2 EE IIT, Kharagpur 7

c. Non-preemptable portions of the interrupt handler routines are kept small and deterministic d. Interrupt handlers are scheduled and executed at appropriate priority e. The maximum time during which interrupts are masked by the OS and by device drivers must be known. f. The maximum time that device drivers use to process an interrupt, and specific IRQ information relating to those device drivers, must be known. g. The interrupt latency (the time from interrupt to task run) must be predictable and compatible with application requirements F. For fast response: a. Run-time overhead is decreased by reducing the unnecessary context switch. b. Important timings such as context switch time, interrupt latency, semaphore get/release latency must be minimum

Point to Ponder: 2
A. What are possible reasons for which execution time for a can be unpredictable? B. Why preemptive multi-tasking is such an important requirement for an RTOS? C. Suppose there are two mutually independent tasks A and B. Let B be the higher priority task. If only task A (B) executes on the processor P, it takes tA ( tB) units of time. If these tasks are run under multi-tasking, neglecting time related to task switching, can the execution time for task A be less than (tA + tB ) ?
B B

Process Management
On a computer system with only one processor, only one task can run at any given time, hence the other tasks must be in some state other than running. The number of other states, the names given to those states and the transition paths between the different states vary with the operating system. A typical state diagram is given in Figure 4 and the various states are described below.

Task States
♦ Running: This is the task which has control of the CPU. It will normally be the task which has the highest current priority of the tasks which are ready to run. ♦ Ready: There may be several tasks in this state. The attributes of the task and the resources required to run it must be available for it to be placed in the 'ready' state. ♦ Waiting: The execution of tasks placed in this state has been suspended because the task requires some resources which is not available or because the task is waiting for some signal from the plant, e.g., input from the analog-to-digital converter, or the task is waiting for the elapse of time. ♦ New: The operating system is aware of the existence of this task, but the task has not been allocated a priority and a context and has not been included into the list of schedulable tasks.

Version 2 EE IIT, Kharagpur 8

♦ Terminated: The operating system has not as yet been made aware of the existence of this task, although it may be resident in the memory of the computer.

admitted new ready

interrupt

exit terminated

running

scheduler dispatch I/O or event completion waiting I/O or event wait

Fig. 37.1 The various states a task can be in during its execution life cycle under an RTOS Task State Transitions When a task is “spawned”, either by the operating system, or another task, it is to be created, which involves loading it into the memory, creating and updating certain OS data structures such as the Task Control Block, necessary for running the task within the multi-tasking environment. During such times the task is in the new state. Once these are over, it enters the ready state where it waits. At this time it is within the view of the scheduler and is considered for execution according to the scheduling policy. A task is made to enter the running state from the ready state by the operating system dispatcher when the scheduler determines the task to be the one to be run according to its scheduling policy. While the task is running, it may execute a normal or abnormal exit according to the program logic, in which case it enters the terminated state and then removed from the view of the OS. Software or hardware interrupts may also occur while the task is running. In such a case, depending on the priority of the interrupt, the current task may be transferred to the ready state and wait for its next time allocation by the scheduler. Finally, a task may need to wait at times during its course of execution, either due to requirements of synchronization with other tasks or for completion of some service such as I/O that it has requested for. During such a time it is in the waiting state. Once the synchronization requirement is fulfilled, or the requested service is completed, it is returned to the ready state to again wait its turn to be scheduled.

Version 2 EE IIT, Kharagpur 9

Point to Ponder: 3
A. What are the situations under which a running task can go to the ready state? B. What are the situations under which a running task can go to the waiting state? C. Suppose there are two mutually independent tasks A and B, which are invoked periodically with periods with B as the higher priority task. If only task A (B) executes on the processor P, it takes tA ( tB) units of time. The periods for A and B are TA and TB, respectively, and tA ≤ TA and tB ≤ TB. If these tasks are to be run under multitasking with negligible task switching time, state any one condition under which task preemption would be essential for the task dead lines to be met.
B B B B

D. In the above case state a condition under which even with preemption, task deadlines cannot be met.

Task Control Functions
RTOSs provide functions to spawn, initialise and activate new tasks. They provide functions to gather information on existing tasks in the system, for task naming, checking of the state of a given task, setting options for task execution such as use of co-processor, specific memory models, as well as task deletion. Deletion often requires special precautions, especially with respect to semaphores, for shared memory tasks.

Task Context
Whenever a task is switched its execution context, represented by the contents of the program counter, stack and registers, is saved by the operating system into a special data structure called a task control block so that the task can be resumed the next time it is scheduled. Similarly the context has to be restored from the task control block when the task state is set to running. The information related to a task stored in the TCB is shown below.

Version 2 EE IIT, Kharagpur 10

Task Control Block
• • • Task ID: the unique identifier for a task Address Space : the address ranges of the data and code blocks of the task loaded in memory including statically and dynamically allocated blocks Task Context: includes the task’s program counter(PC) , the CPU registers and (optionally) floating-point registers, a stack for dynamic variables and function calls, the stack pointer (SP), I/O device assignments, a delay timer, a time-slice timer and kernel control structures Task Parameters : includes task type, event list Scheduling Information : priority level, relative deadline, period, state Synchronization Information : semaphores, pipes, mailboxes, message queues, file handles etc. Parent and Child Tasks

• • • •

Task Scheduling and Dispatch
The basic purpose of task scheduling and dispatch in a real-time multi-tasking OS is to ensure that each task gets access to the CPU and other system resources in a manner that is necessary for successful and timely completion of all computation in the system. Secondly, it is desired that this is done efficiently from the point of view of resource utilisation as well as with correct synchronisation and protection of data and code for individual tasks against incorrect interference. Various task scheduling and dispatch models are in use to achieve the above. The appropriateness of a particular model depends on the application features. The major main task scheduling and dispatch models are described below.

Cyclic Executive
This is the simplest of the models in which all the computing tasks are required to be run periodically in cycles. The computing sequence is static and therefore, a monolithic program called the Cyclic Executive runs the tasks in the required order in a loop. At times the execution sequence may also be determined in accordance with models such as an FSM. The execution sequences and the times allocated to each task are determined a priori and are not expected to vary significantly at run time. Such systems are often employed for controllers of industrial machines such as Programmable Logic Controllers that perform fixed set of tasks in fixed orders defined by the user. These systems have the advantage of being simple to develop and configure as well as faster than some of the more complex systems since task context switching is faster and less frequent. They are however suitable for static computing environments only and are custom developed using low level programming languages for specific hardware platforms.

Coroutines
In this model of cooperative multitasking the set of tasks are distributed over a set of processes, called coroutines. These tasks mutually exchange program control rather than relinquish it to the Version 2 EE IIT, Kharagpur 11

operating system. Thus each task transfers control to one of the others after saving its data and control state. Note that the responsibility of scheduling, that is deciding which task gets control of the processor at a given time is left to the programmer, rather than the operating system. The task which is transferring control is often left in the waiting or blocked state. This model has now been adapted to a different form in Multithreading.

Interrupts
In many cases task scheduling and dispatch needs to be made responsive to external signals or timing signals. In other cases running tasks may not be assumed to be transferring control to the dispatcher on their own. In such cases the facility of interrupts provided on all processors can be used for task switching. The various tasks in the system can be switched either by hardware or software interrupts. The interrupt handling routine would then transfer control to the task dispatcher. Interrupts through hardware may occur periodically, such as from a clock, or asynchronously by external devices. Interrupts can also occur by execution of special software instructions written in the code, or due to processing exceptions such as divide by zero errors. Interrupt-only systems are special case of foreground/background systems, described below, which are widely used in embedded systems.

Foreground / Background
Typically, embedded and real-time applications involve a set of tasks, some of which are periodic and must be finished within deadlines. Others may be sporadic and may not have such deadlines associated with them. Foreground/background systems are common and simple solutions for such applications. Such systems involve a set of interrupt driven or real-time tasks called the foreground and a collection of non-interrupt driven tasks called the background. The foreground tasks run according to some real-time priority scheduling policy. The background tasks are preemptable by any foreground task.

Real Time Operating Systems
This is the most complex model for real-time multi-tasking. The major features that distinguish it from the other ones described above are the following. 1. The explicit implementation of a scheduling policy in the form of a scheduler module. The scheduler is itself a task which executes every time an internal or external interrupt occurs and computes the decision on making state transitions for every application task in the system that has been spawned and has not yet been terminated. It computes this decision based on the current priority level of the tasks, the availability of the various resources of the system etc. The scheduler also computes the current priority levels of the tasks based on various factors such as deadlines, computational dependencies, waiting times etc. 2. Based on the decisions of the scheduler, the dispatcher actually effects the state transition of the tasks by a. saving the computational state or context of the currently executing task from the hardware environment. b. enabling the next task to run by loading the process context into the hardware environment. Version 2 EE IIT, Kharagpur 12

It is also the responsibility of the dispatcher to make the short-term decisions in response to, e.g., interrupts from an input/output device or from the real-time clock. The dispatcher/scheduler has two entry conditions: 1. The real-time clock interrupt and any interrupt which signals the completion of an input/output request 2. A task suspension due to a task delaying, completing or requesting an input/output transfer. In response to the first condition the scheduler searches for work starting with the highest priority task and checking each task in priority order. Thus if tasks with a high repetition rate are given a high priority they will be treated as if they were clock-level tasks, i.e., they will be run first during each system clock period. In response to the second condition a search for work is started at the task with the next lowest priority to the task which has just been running. There cannot be another higher priority task ready to run since a higher priority task becoming ready always preempts a lower priority running task.

task scheduler/ dispatcher System Services I/O, interrupt API calls task management memory management I/O management initialization

kernel external interrupts clock/timer interrupts i/o interrupts application tasks

Fig. 37.2 Structure of an RTOS Kernel The typical structure of an RTOS kernel showing the interaction between the System and the Application tasks.

Version 2 EE IIT, Kharagpur 13

Point to Ponder: 4
A. In what ways threads are similar to coroutines? In what ways are they different? B. Under what situations is a cyclic executive adequate? When is it not and why not? C. In what ways are interrupt service routines different from other tasks? D. Can you give an example of a practical embedded application, for which an Foreground/ Background model of multi-tasking is adequate? E. Under what situations is an full-featured RTOS an appropriate choice? Can you illustrate your answer with the example of a practical embedded application?

Priority Levels in a typical Real-Time Operating System
To be able to ensure that response to every event is generated by executing tasks within specific deadlines, it is essential to allocate the CPU and other computational resources to various tasks in accordance with their priorities. The priority of a process may be fixed or static. It may be calculated based on their computing time requirements, frequency of invocations or deadlines. However, it has been established that policies that allow task priorities to be adjusted dynamically are more efficient. The priority will depend on how quickly a task will have to respond to a particular event. An event may be some activity of the process or may be the elapsing of a specified amount of time. The set of tasks can often be categorised into three broad levels of priority as shown below. Tasks belonging to the same category are typically scheduled and dispatched using similar mechanisms. Interrupt Level: Tasks at this level require very fast response measured in milliseconds and occur very frequently. There is no scheduler at this level, since immediate execution follows an interrupt. Examples of this task include the real-time clock task. Obviously, to meet the deadlines of the other tasks in the system, the context switching and processing time requirements for these tasks are to be kept at the bare minimum level and must be highly predictable to make the whole system behaviour predictable. To ensure the former, often, all Interrupt Service Routines (ISRs) run in special common and fixed contexts, such as common stacks. To ensure the latter the interrupt service routines are sometimes divided into two parts. The first part executes immediately, spawns a new task for the remaining processing and returns to kernel. This new task gets executed as per the scheduling policy in course of time. The price to pay for fast execution of ISRs is often several constraints on the programming of ISRs which lacks many flexibilities compared to the programming of tasks. Priorities among tasks at this level are generally ensured through the hardware and software interrupt priority management system of the processor. There may also exist interrupt controllers that masks interrupts of lower priority in the presence of a higher priority one. The system clock and watchdog timers associated with them are tasks that execute at interrupt level. The dispatcher for the next level of priority is also a task at this level. The frequency of execution of this task depends on the frequency or period of the highest priority clock-level task. Hard Real-Time Level: At this level are the tasks which are periodic, such as the sampling and control tasks, and tasks which require accurate timing. The scheduling of these tasks is carried out based on the real-time system clock. A system clock device consists of a counter, a timer queue and an interrupt handler. Content of counter gives the current time, timer queue has Version 2 EE IIT, Kharagpur 14

pending timers associated with the clock device. Each system clock interrupt is known as a tick and represents the smallest time interval known to the system. Since most programs in real-time applications make use of time, virtual software clocks and delays can also be created by tasks and associated with the system clock device. The system clock device raises interrupts periodically and the kernel updates the software clock according to current time. Also every few clock cycles a new task gets dispatched according to the scheduling policy adopted. The lowest priority task at this level is the base level scheduler. Thus if at a clock level interrupt, the clock level scheduler finds no request for higher priority clock level tasks pending, the base level scheduler is dispatched. Soft/Non-Real Time Level: Tasks at this level are of soft or non-real-time in that they either have no deadlines to meet or are allowed a wide margin of error in their timing. These are therefore taken to be of low priority and executed only when no request for a higher priority task is pending. Tasks at this level may be allocated priorities or may all run at a single priority level - that of the base level scheduler in a round robin fashion. These tasks are typically initiated on demand rather that at some predetermined time interval. The demand may be a user input from a keypad, some process event such as the arrival of a packet or some particular requirement of the data being processed. Note that, since the base level scheduler is the lowest priority clock level task, the priorities of all base level tasks are lower than those at clock levels. Each of the above priority levels can support multiple task priorities. For an RTOS to be compliant with the RT-POSIX standard, number of priority levels supported must be at least 32. Among commercial RTOS, the priority levels supported can be as low as 8 for Windows CE or 256 for VxWorks. For scheduling of equal priority threads, FIFO or Round-Robin policy is generally applied. Thread priorities be changed at run-time.

Task Scheduling Management
Advanced multi-tasking RTOSs mostly use preemptive priority scheduling. These support more than one scheduling policy and often allow the user to set parameters associated with such policies, such as the time-slice in Round Robin scheduling where each task in the task queue is scheduled for up to a maximum time, set by the time-slice parameter, in a Round Robin manner. Task priorities can also be set. Hundred of priority levels are commonly available for scheduling. Specific tasks can also be indicated to be nonpremeptive.

Point to Ponder: 5
A. Give example of hard, soft and non-real-time computing tasks. In what ways are they different? B. Can interrupt level priorities be seen as very high priority levels? C. What happens if the user sets priorities in a manner that deadline would be violated for other tasks?

Lesson Summary
In this lesson, we have dealt with the following topics. Version 2 EE IIT, Kharagpur 15

A. Basic Purpose of an Operating System B. The features that characterise a real-time operating system C. The major concepts involved in the Process Management function of the RTOS a. Task Dispatching b. Preemptive priority scheduling D. Basic Styles of implementing the Process Management in RTOS’s of varying complexities

Version 2 EE IIT, Kharagpur 16

Answers, Remarks and Hints to Points to Ponder Point to Ponder: 1
A. What are the factors on which the execution time of a task depends on? Ans: The execution time of a task can be divided into two types of activities, namely, CPU time and I/O time. In a multi-tasking operating system a third kind of time is added to this, namely the time spent by the task in waiting for the resources needed, that is CPU, or I/O. This may depend on a variety of factors, such as the other tasks running in the environment, priority, scheduling policies etc. B. While a task is executing, is the CPU continuously busy? Ans: The CPU is always doing something except during the times when it is fetching address or data from memory or devices. But since it may require sometime for the device to send the data, during that time, the CPU may do tasks other than the one for which it requested the device. In this sense the CPU may not be continuously busy. C. Suppose there are two mutually independent tasks A and B. If only task A (B) executes on the processor P, it takes tA ( tB) units of time. If these tasks are fired simultaneously and run under multi-tasking, neglecting time related to task switching, would the execution time for both tasks taken together be greater than, equal to or less than (tA + tB ) ?
B B

Ans: That depends on the scheduling policy. If the policy is non-preemptive, it will take time equal to (tA + tB ). If it is preemptive, however, it is expected to take less time than (tA + tB ).
B B

Point to Ponder: 2
A. What are possible reasons for which execution time for a task can be unpredictable? Ans: Basically due to variations in its execution environment. This includes factors such as the other tasks which are executing at the same time, there priorities related to the task in question, the nature and frequency of interrupts coming from the environment etc. Note that, even apart from these environmental factors, the execution time of a task depends on the input data set used for the run. However, this cannot be termed “unpredictable”. B. Why preemptive multi-tasking is such an important requirement for an RTOS? Ans: The first reason is that, without it, the concept of priority cannot be implemented properly. Thus without preemption, a task, that may have started when no other higher priority task was present, can block such higher priority tasks for long times, thus violating the principle of priority scheduling. Secondly, with a given computing speed the CPU utilization that can be realised with preemption cannot be utilised without it. Even if limited preemption, (namely that, only a task waiting for I/O is preempted) is used, a given set of tasks that is schedulable with respect to their deadlines under preemptive scheduling, may not be so without premption. It is for these reasons that it is an indispensable feature for an RTOS. Version 2 EE IIT, Kharagpur 17

C. Suppose there are two mutually independent tasks A and B. Let B be the higher priority task. If only task A (B) executes on the processor P, it takes tA ( tB) units of time. If these tasks are run under multi-tasking, neglecting time related to task switching, can the execution time for task A be less than (tA + tB ) ?
B B

Ans: Yes, because, while task B is waiting for I/O, it would be removed to a waiting queue and task A would be scheduled till the time the interrupt for I/O completion is received and task B is put into a ready queue.

Point to Ponder: 3
A. What are the situations under which a running task can go to the ready state? Ans: From the new state, after it is initialised. From the waiting or blocked state, after I/O completion. From the running state, after its allocated time-slice is spent. B. What are the situations under which a running task can go to the waiting state? Ans: If it is blocked on I/O or due to synchronization requirement with another task. C. Suppose there are two mutually independent tasks A and B, which are invoked periodically with periods TA (TB). If only task A (B) executes on the processor P, it takes tA (tB) units of time. Naturally, tA ≤ TA and tB ≤ TB. If these tasks are to be run under multi-tasking with negligible task switching time, state any one condition under which task preemption would be essential for the task dead lines to be met.
B B B B

Ans: tB ≤ TA.
B

D. In the above case, let B be the higher priority task which can preempt the execution of task A whenever it is ready. State a condition under which even with preemption, task deadlines cannot be met. Ans: ΤΑ ≥ ⎡ΤΑ /tB ⎤ tB + t A .

Point to Ponder: 4
A. State one factor in respect of which, threads are similar to coroutines. State one factor in respect of which they are different. Ans: In the sense that they do not involve process management overheads on the OS, such as creation and management of TCB, run-time synchronization management etc. They are different in the sense that, there is RTOS support for thread scheduling. Co-routines are entirely managed by the programmer. B. State one situation under which a cyclic executive is adequate? State one situation under which it is not and why not? Version 2 EE IIT, Kharagpur 18

Ans: When task execution times are predictable and the tasks can be run in a given fixed order without violating their deadlines, a cyclic executive is adequate. If the above conditions are not met, for example, if tasks are invoked by the environment asynchronously, a cyclic executive is not the choice. C. State one way in which are interrupt service routines different from other tasks? Ans: They are different in the sense that they receive an immediate response from the processor, if they are sensed (i.e. they are not masked). This is by the hardware design of the processor. Tasks on the other hand receive response from the CPU only when they are appropriate to be run under the scheduling policy. D. Can you give an example of a practical embedded application, for which an Foreground / Background model of multi-tasking is adequate? Ans: Consider an embedded CNC machine controller. The tasks in the machine can typically be divided into two categories. In the first category are real-time supervisory control tasks that generate the references for the two dimensional motion of the job as well as for the spindle speed. These references are generated by interpolating the cutting profiles programmed in the machine. On the other hand there are other tasks which are soft/non-real time, such as servicing the operator input console and updating the display. We do not need a full features RTOS here since the nature of tasks and the task loading level is fairly deterministic and predictable. E. Under what situations is a full-featured RTOS an appropriate choice? Can you illustrate your answer with the example of a practical embedded application? Ans: A full featured RTOS is the choice for a complex embedded system where the task loading can vary a lot and can consist of hard real-time tasks of various periods and deadlines. A typical application area is avionics, such as, say, for a fighter aircraft. The task loading levels can vary widely depending on mission profiles as well as emergency situations. Many of these tasks are mission critical. Also the expertise level of people building these is very high and therefore the complexity of such systems can be handled.

Point to Ponder: 5
A. Give example of hard, soft and non-real-time computing tasks. In what ways are they different? Ans: Consider a digital camera. Among the various tasks that are performed by the embedded processor, the shutter control functions which essentially compute the necessary exposure based on the light conditions and opens and closes the shutter is a hard real-time task. If the task does not execute in time, the exposure would change. On the other hand, the task that displays the picture on the display, as one reviews the pictures already taken, is a non-real-time task, because there is no fixed deadline for the task. Note that this task also must execute reasonably fast (otherwise the camera would not sell). On the other hand, the auto-focus task, on such a camera which allows one to zoom on a target may be classified as soft-real-time because if the camera does not focus exactly before the shutter is pressed, the Version 2 EE IIT, Kharagpur 19

performance of the camera would degrade in terms of photographic quality in some cases. Thus, the categorisation of the task among the three classes is actually intimately related to the criteria of performance that is decided for the system. B. Can interrupt level priorities be seen as a task at real-time level with very high priority? Ans: Interrupts differ from high priority real-time tasks, although in many cases a part of an interrupt service routine may actually be spawned as a real-time high priority task. Firstly, the interrupt response is built in into the hardware, whether it is generated through a hardware pin or a software instruction, and is automatic, once the interrupt is sensed. It is not computed using a scheduling policy as is the case for an RT task. Thus an interrupt does not have a deadline associated with it like an RT task. It may or may not be periodic. C. What happens if the user sets priorities in a manner that deadline would be violated for other tasks? Ans: RT systems generally do not admit of “users” who can set priorities of tasks at runtime. However, an application programmer can write code that spawns new tasks at specified priorities. If these are not set properly deadlines would indeed be failed and the system would be considered to have been implemented according to specifications. In fact, it is for this reasons that embedded systems are tested extensively to ascertain that the all deadlines can be met under all kinds of task scenarios.

Version 2 EE IIT, Kharagpur 20

Module 8
Industrial Embedded and Communication Systems
Version 2 EE IIT, Kharagpur 1

Lesson 38
Networking of Field Devices via Fieldbus
Version 2 EE IIT, Kharagpur 2

Lesson Objectives
• • • • To motivate a field level networked digital communication architecture for implementation of distributed plant wide control To describe the Fieldbus network protocol To describe the basic computation and communication architecture for Fieldbus devices To explain issues related to time synchronization, interoperabilty, communication efficiency etc. in the Fieldbus network.

1. Introduction
Embedded electronics technology has given rise to significant rise in the number of automatic devices for industrial data acquisition, transmission, monitoring, diagnostics, control and supervision. Each of these devices is configurable and capable of two way communication with other devices. Effective use of their capabilities can only be enabled by reliable and high speed communication architecture for extensive and rapid information exchange among automation devices for coordination and control. Below we introduce some of the major motivations that led to major users and suppliers from the U.S., Japan and Europe coming together to establish the Fieldbus Foundation in 1994. Their objective has been to develop a worldwide, unified specification of "Fieldbus", a network communication architecture for field devices for process control and manufacturing automation.

Motivations for the Fieldbus
Among the major motivations for the Fieldbus are the following. • Replacement of analog and digital (serial) point-to-point communication technology with much superior digital communication network for high speed ubiquitous and reliable communication within a harsh industrial environment. Enhanced data availability from smart field bus devices needed for advanced automation functions such as control, monitoring, supervision etc. Easy configurability and interoperability of system components leading to an easily installable, maintainable and upgradeable open system that leverages the computing and networking hardware and software solutions

• •

In industrial automation systems, the field signals have been traditionally transmitted to the control room using point-to-point communication methods that employ analog technologies such as the 4-20 mA current loop or, more recently, digital ones such as the RS-422 or RS-485. The main disadvantages of this are the highly increased cost of cabling due to the need for a separate pair of wires for each device connected to the mainframe. Apart from this, with 4-20 mA analog current loop, signals can be transmitted only in one direction. With the need for more complex monitoring and control of a process plant, installation and maintenance of these point to point communication media and their signal integrity become more and more difficult. As an alternative the network communication architecture presents an attractive option. Firstly the cabling requirements are marginally increased as more and more devices are added to the Version 2 EE IIT, Kharagpur 3

network. Secondly, a vast array of high speed networking technologies is available at attractive costs from the computer market. Thirdly, with the addition of intelligent devices, such a system enables advanced monitoring supervision and control, leading to improvements in productivity, quality and reliability of industrial operations.
Marsh -alling box

Junction Box

Wiring Duct

Device Controller I/o card

A JB Field Bus Device A

Wire Duct Local Controller HMI

P

Fig. 38.1 Wiring system for conventional point-to-point communication systems and the Fieldbus Fieldbus is a standard for Local Area Network (LAN) of industrial automation field devices that enables them to intercommunicate. Typical Fieldbus devices are sensors, actuators, controllers of various types, such as PLCs, and DCS, and other computer systems such as human-machine interfaces, process management servers etc. It includes standards for the network protocol as well as standards for the devices on the network. Fieldbus allows many input and output variables to be transmitted on the same medium such as, a pair of metallic wires, optical fibre or even radio, using standard digital communication technologies such as baseband time-division multiplexing or frequency division multiplexing. Thus sensors transmit the measured signal values as well as other diagnostic information; the controllers compute the control signals based on these and transmit them to actuators. Further, advanced features such as process monitoring can be carried out leading to increased fault tolerance. Online process auto-tuning can be performed leading to optimized performance of control loops. Table 1 compares some of the key features of 4-20mA and Fieldbus technology. It should be mentioned that Fieldbus becomes cost-effective only beyond a certain scale of operations.

Version 2 EE IIT, Kharagpur 4

Table 38.1 Comparison of Fieldbus with 4-20mA current loop Item No. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. Specification No. of devices per wire Qty. of data/variables per device Control functions in field Device Failure Notification Signal degradation over wire Power distribution over wire Interchangeability of field devices Maximum run-length Failure diagnosis Intrinsic safety Sampling delay 4-20mA 1 1 No Minimal (O/C, S/C) Possible Yes Yes 2Km Technician reqd. With barriers Vendor defined Fieldbus 32 Up to thousands Yes Yes, detailed None Yes Yes (with some restrictions) 1.9Km (5.7 Km with repeaters) Operator informed at console With barriers User defined (within limits)

Fieldbus technology was designed for geographically distributed harsh environment of process control applications. Also, it was conceived that there would be frequent changes in the installations. To meet these requirements the protocol includes the following aspects which are not necessarily found in other Protocols: • • Control algorithms may be in field-mounted Devices, central controlled or a combination of both. The End User does not have to be concerned with Device numerical address allocation. The Protocol handles this task, so 'plug and play' services are available for commissioning, modification and replacement. Devices do not have to be 'configured' before they are attached to the network. Device Definition and Function Blocks create a standard vendor-independent device interface for each device type which, in turn, facilitate installation, commissioning and upgradation of multi-vendor applications. The Physical Layer of the Protocol was designed from the outset to cope with installed cables and flammable atmospheres (hazardous areas). Both precise cyclic updates as well as acyclic and sporadic communications are catered for within the Protocol. Each variable transmitted on the Fieldbus carries with it tags indicating the current health of the source. Using this information, recipient Devices can take appropriate action immediately (for example switch to Manual, Off-line, etc.).

• •

• • •

Version 2 EE IIT, Kharagpur 5

Point to Ponder: 1
A. How exactly does a network lead to reduced cabling over a point-to-point communication topology? B. Can you cite three reasons why the reduction in cabling is considered a significant advantage of the Fieldbus? C. Can you state two advantages of interoperability? D. For what kind of factories is a Fieldbus implementation for plant automation justified?

2. Fieldbus Topology
As shown in Figure 38.1, Fieldbus generally uses one of the two topologies - Bus and Tree. With the Bus Topology, devices are connected to the network 'back-bone'. Either through a 'Drop Cable' Device, or are directly connected to the Bus by a 'Splice' connection. The Tree arrangement is used where a number of Devices share a similar location remote from the equipment room. A junction box, installed at the geographic center of gravity of the Devices, communicates with each Device is connected to it via a cable. In general Fieldbuses can use a combination of both topologies. Thus, trees can be hung from network buses. Tree Bus with Drop Cables

Remote I/O Drop cable

Device Flow Positioner Thermo meter Couple

Remote I/O

Fig. 38.2 Tree and Bus Structures for Fieldbus

3. Architecture of the Fieldbus
The Open Systems Interconnect (OSI) model published by the International Standards Organization is a well known definition of network communications based on seven generic layers. It defines seven generic 'Layers' required by a communication standard capable of supporting vast networks. The first two layers, namely the Physical and the Data Link layers incorporate the technologies to realize a reliable, relatively error free and high speed communication channel among the communicating devices. It provides support for all standard and medium dependent functions for

Version 2 EE IIT, Kharagpur 6

physical communication. DLL actually manages the basic communication protocol as well as error control set up by higher layers. In Fieldbus, since the communication takes place over a fixed network routing and transport layers are made redundant. Moreover, in an industrial control environment, the network software entities or processes are also generally invariant. Under such a situation, requirements of the session and the presentation layers are also minimized. Therefore, the third, fourth, fifth and sixth layers of the ISO protocols have been omitted in the Field bus protocol. In fact the requirements of the omitted layers, although limited, have been included within the Fieldbus Application Layer (FAL) (7), which is sub-divided into two sub-layers, namely the Fieldbus Message Sub-layer (FMS) Fieldbus Message Sub-layer (FMS) that builds up a message data structure for communication as per requirements of user layer and includes the roles of the session and presentation layers of the ISO-OSI model , and the Field Access Sub-layer (FAS) that manages the functionality of the networking and transport layers to the extent needed and provides a virtual communication channel. Thus, the Foundation Fieldbus utilizes only three ISO model Layers (1, 2 and 7), plus an additional Layer referred to as the User Layer (8). In the Fieldbus standard, the User Layer (8) is also included in the specification. In this it differs from other communication standards. A typical function of the User Layer is to define control tasks for a process plant. These are achieved through abstract software units called Function Blocks. Defining the User Layer functionality in terms of the open and published standards of Function Blocks enables interoperability of devices from different vendors. This is because any two devices that implement the standard abstract function block interface would interoperate, irrespective of their internal implementations. Fieldbus Foundation has standardised a range of Function Block communications interfaces. The content of a Function Block is not standardised. For example: Company A and Company B may both supply PID control algorithms within their products. The Fieldbus Foundation specification dictates how each vendor's PID Function Block shall communicate Set-point, Controlled Variable, P, I &D constants etc., but not how the Function Block's internal algorithms would be realised. The fieldbus protocol structure is shown alongside with that of the ISO-OSI model in Fig. 2. Below we discuss each of the above layers of the Fieldbus in more detail.

Version 2 EE IIT, Kharagpur 7

User Layer 7. Application Layer 6. Presentation Layer 5. Session Layer 4. Transport Layer 3. Network Layer 2. Data Link Layer 1. Physical Layer

SM System Management

FB Function Block

DD

FAL (FMS) Message Sub layer Network Management FAL (FAS) Access Sub layer Omit Layer 3 thru 6 DLL Data Link layer PHY Physical layer

Fig. 38.3 Fieldbus Network Architecture vis-à-vis OSI

Point to Ponder: 2
A. Why are network and transport layers absent in a Fieldbus network? B. Why are session and presentation layers absent in a Fieldbus network?

4. The Physical Layer
Fieldbus allows options for three types of communication media at this layer, namely, Wire, Fiber-optic and Radio. The Physical Layer is sub-divided into an upper section (the Media Independent Sub-Layer MIS) and a lower section which is media specific. The MIS ensures that the selected Media interfaces in a consistent way with the Data Link Layer (2), regardless of the media used. The lower sections define the communications mechanism and media. For example, for wire medium they describe signal amplitudes, communication rate, waveform, wire types, etc. An area-wide network can be implemented through the compartmentalization of the bus system in the bus segments that can be connected over repeaters. Standard-transmission rates can be in the range of from about 10KBaud 10 MBaud. The topology of the single bus segment is the line structure (up to 1200 m) with short drop cables (<0.3m). Transmission distances to 12 km are possible by electrical configuration and to 23.8 km with optical configuration. The distances are dependent on the transmission rate. With the help of repeaters, a tree structure can also be constructed as shown:

Version 2 EE IIT, Kharagpur 8

1

2

3

4

5

6

7

8

9 Segment

10

11 Node

12

13 Repeater

14

Fig. 38.4 Multi-bus segment Fieldbus network topology. The maximum number of nodes per bus segment amounts to 32. More lines can connected under one another through performance enhancements (repeaters) where by it is noted that each repeater counts as a node. In total a maximum of 128 nodes are connectable (over all bus segments).

Point to Ponder: 3
A. Why are repeaters required for extending the bus in a network? B. Why does the maximum bus length depend on the transmission rate?

6. The Data Link Layer
As the medium of transmission is a bus network, all device communications take place over the same physical medium. A mechanism is therefore necessary to ensure that it is shared effectively without collisions, i.e., when one device transmits none other does. The Fieldbus Data Link Layer protocol is a hybrid protocol that is capable of supporting both scheduled and asynchronous transfers. Its maximum packet size is 255 bytes. It defines three types of data link layer entities, a Link Master(LM), a Basic Device(BD), and a bridge. Link master devices are capable of assuming the role of the bus master, called the link active scheduler (LAS). At any point of time only one of the LM devices act as the LAS. This is depicted in Figure 38.3.

Version 2 EE IIT, Kharagpur 9

High Speed Ethernet Idle LAS LD Idle

LAS LM

BD

BD

BD

LAS LM

Fig. 38.5 Link Active Scheduler, Link Masters and Basic Devices for a Fieldbus implemented on a High Speed Ethernet Basic devices are those devices not capable of becoming the LAS. They receive and send published data, and they receive and use tokens. When they hold the token, they are capable of initiating communications with all devices on the network. Bridge devices connect link segments together. Bridged networks are configured into a spanning tree in which there is a single root link segment and a series of downstream link segments. Bridges interconnect the link segments. Each bridge may have a single upstream port (in the direction of the root) and multiple downstream ports (away from the root). The root port behaves as a basic device and the downstream ports are each the LAS for their downstream link. Bridges are responsible for republishing scheduled transfers and forwarding all other traffic. Configured republishing and forwarding tables identify the packets that are to receive and republish or forward. Bridges are also responsible for synchronizing time messages received on their root port before regenerating them on their downstream ports.

6.1 The Link Active Scheduler (LAS)
One of the devices connected to the Fieldbus acts as the Link Active Scheduler (LAS). This decides which Device transmits next and for how long, thereby avoiding the collision of messages on the Bus. The LAS is responsible for the following list of tasks. 1. It detects the connection and disconnection of devices to the network, in order to maintain a "Live List" of functional devices and ensure they receive the "Right to transmit" when appropriate. Redundant LAS's maintain their own Live Lists in readiness to take over when the on-line LAS fails 2. It distributes time on the bus that can be used for scheduling and time stamping. 3. It polls device buffers for data according to a predefined schedule. This capability is used to support publisher/subscriber virtual communication relationships. 4. It distributes a token to devices in its live list that they can use for asynchronous transfers. This capability is used to support client/server and report distribution virtual communication relationships.

Version 2 EE IIT, Kharagpur 10

The LAS controls all cyclic data transmissions in this manner. In free time the LAS passes a message called the Pass Token (PT) to each Device in turn allowing them to use this idle period. As mentioned before, the Link Active Scheduler (LAS) controls communications traffic on the Fieldbus. This is also called "Bus master function". The active LAS grants a "right to transmit" to each device on Fieldbus in a pre-defined manner. Devices other than LAS can communicate only when they have the "right to transmit". There are two ways of granting a "right to transmit". One is a polling method, which grants a right to transmit in sequence to each device. Another is a time slot method, which grants a right to transmit at a fixed time interval. The LAS uses these two methods combined to meet the requirements of precise cyclic updates and unscheduled traffic, for example, alarm reporting.

6.1 Cyclic Communication
Typically, cyclic communications in industrial operations involving input output operations related to process control loops or PLC scan cycles. Such communications must be performed at precise update rates. The LAS meets the requirement of precise cyclic updates of variables by issuing a "Compel Data" message (called the CD Token), to each source of data according to a fixed schedule. On receiving the CD, the addressed device transmits the current data on the bus. This message contains a reference to the source of the data. Any other device on the bus requiring the data takes a copy for its own use, for example an HMI or a control loop. Note that only one transmission is required to satisfy many destinations. The device transmitting the data is referred to as the "Publisher" and those who take copies are called "Subscribers". The publisher may not know which devices are subscribers. The publisher's data is referred to as a Data Transfer Process Data Unit, or DT for short.

Host

Execution of PID Control and Communication

HMI PID 101 H1 AO 101

AI 101

Fig. 38.6 Communication within a Control Loop If a control loop requires a measured variable to be updated on a cyclic basis, the LAS instructs the source of the signal to transmit the variable by sending a special message called the Compel Data (CD) token. On receiving this message, the source transmits the variable on the bus. All devices on the bus receive the message, but only those with a use for the information take a copy. Version 2 EE IIT, Kharagpur 11

In Figure 38.6, the Process Variable (PV) sensor transmits the measured variable when it receives the CD token. This is referred to as 'Publishing' the data. The control algorithm in the control valve copies it, as it is a Subscriber to this information. The HMI may also copy it for display and archiving purposes, but only one transmission of the PV is required.

6.2. Acyclic/Unscheduled Communication
Apart from cyclic communications, requirements for acyclic communications arise to handle sporadic process related events, such as, • • • • • Alarm Operator Data Update Trend Data Update Set Point changes Controller Tuning

Once the requirements for cyclic data transmission have been met, the LAS will issue a Pass Token (PT) to each device in turn, thereby allowing them access to the bus to transmit data (a DT) or request data from another device, utilizing the bus up to an allocated time limit.

6.3 Macro Cycle and Elementary Cycle
A basic requirement of process control applications is that precise cyclic updates of process variables should be possible, to ensure good quality continuous control. Generally the number of such tasks in the system remains more or less fixed. Apart from these, communication tasks related to sporadic events, such as alarm reporting and operator changes of set points, must be scheduled. The LAS therefore organizes its overall schedule communication tasks in the system in “Macro Cycles”. The duration of each Macro Cycle is further subdivided into a number of “Elementary Cycles”. This is shown in Figure 38.7.
Macro Cycle

Elementary Cycle

Elementary Cycle

Fig. 38.7 Macro Cycles and Elementary Cycles Each EC within an MC begins with the set of periodic tasks that is to be scheduled within that EC according to its update time period. The EC is chosen to be of such a duration that even after processing of the periodic tasks some time is left for servicing aperiodic tasks, should it be necessary, due to the occurrence of some event in the system.

Version 2 EE IIT, Kharagpur 12

1 Sec. Macro Cycle
CD.1 DT.1 CD.2 DT.2 PT.n DT.n CD.1 DT.1 PTn+1 DTn+1 PTn+2 DTn+2 CD.1 DT.1

Cyclic Elementary Cycle 1

Acyclic

Cyclic

Acyclic

Time

Elementary Cycle 2 Fig. 38.8 An example task schedule

This is shown in Figure 38.8 in the case of a simple example of a system containing two devices requiring cyclic updates. The update requirements of the two devices are 1 sec. and 0.5 sec. respectively. The LAS sets the EC period as equal to the shortest update time requirement (0.5 Second. in this case). Similarly, the longest update time sets the MC period (1 Second. in this case). The CD for Device 1 is generated at the beginning of each Elementary Cycle and the CD for Device 2 at the second time slot of alternate Elementary Cycles. In the 'free time' in each Elementary Cycle the LAS transmits PT's to devices on the Fieldbus segment in turn, allowing them to transmit unscheduled information. This is the unscheduled portion of the Elementary Cycle. There may be insufficient free time for all Devices to receive a PT before the end of the Elementary Cycle. In this case the LAS continues from where it left off in subsequent cycles. Note that the time available for unscheduled traffic varies from one Elementary Cycle to another. For example in the first EC of an MC in the example, both periodic updates takes place, while in the second EC only one does, since the update requirement of device 2 is lower. Also, the CD's requiring the shortest update intervals are dealt with first in each Elementary Cycle. Thereby ensuring the interval between subsequent updates remains constant.

Point to Ponder: 4
A. How does the LAS know how much time to give to which process and in what sequence? B. Is the proposed medium access strategy adequate for industrial automation? C. Can you think of an alternative strategy for network communication?

7. The Application Layer
The objective of the Application Layer is to convert data and requests for services coming from the User Application (Layer 8), into demands on the communication system in the Layers below, and to provide the reverse service for received messages. Thus the application layer abstracts the technical details of the network from the user layer which can view the network devices to which communication is needed as if they are connected by virtual point to point communication channels. The Application Layer is subdivided into two sublayers namely the Fieldbus Access Sublayer (FAS) and the Fieldbus Message Sublayer (FMS). These are described below.

7.1 Fieldbus Access Sublayer
The FAS sits in-between the FMS and the DLL. The FAS provides three fundamental kinds of Version 2 EE IIT, Kharagpur 13

communications. The services offered by the higher layers such as the FMS are realized by the FAS using one of these modes of communication. They are described below.

7.1.1. One-to-one Bi-directional (QUB)
QUB is used for the communication between a device, which requests data on Fieldbus, and a device which provides the data. A typical example of such a communication is screen display updates and the change of setting of the function block parameter, etc. through an operator's station. QUB is initiated by a device (client) requesting read/write of parameters, and is terminated when another device (server) returns a response. Therefore, the communication is of the bi-directional, and confirmed, that is with acknowledgement from the server.

7.1.2. One-to-one Unidirectional 1 (BNU)
This type of communication is used for the distribution of data, which is generated in one device (publisher) and is transmitted to one or more devices (subscribers). The publishing application writes the data into a distributed network buffer. The network is responsible for copying the data to corresponding network buffers in subscriber devices. Subscribing applications subscribe to published data asynchronously by opening buffers for the receipt of the published data and identifying the associated publisher. A typical example is a pressure transmitter sending measurement data as a Process Variable (PV), and a valve positioner receiving it and using it to modulate a valve. Unlike QUB, BNU is initiated cyclically according to schedule, not by a request for data. Neither does it involve a response from the server. This is an unconfirmed communication service in single direction. BNU uses a connection-oriented service in the data link layer.

7.1.3. One-to-one Unidirectional 2 (QUU)
With QUU, one device on Fieldbus generates data, and interested recipients take a copy. The transmitter is referred to as the 'Source' and the recipients as 'Sinks'. This is typically for multicasting event reports and trend reports. Unlike published data, reports are sent to preconfigured group addresses when the bus is not scheduled for the transfer of published data. These virtual communication relationships used connectionless transfers. Note that QUU differs from BNU in that communication is unscheduled and new data does not overwrite older data in the recipient devices.

7.2 The Fieldbus Message Sublayer (FMS)
The FMS acts as the interface between the User Layer and FAS. There is a logical framework called Virtual Field Device (VFD), which manages various functions and parameters at the user layer. A Fieldbus Device must have at least two VFD's, one for administering the network, the other for the control of the system or function blocks. The former has the parameters related to setting up the communication, the latter has the parameters related to Function Blocks defined by user layer and required by the control application. Version 2 EE IIT, Kharagpur 14

The process control oriented VFD in a Fieldbus device is its Function Block Application Process (FBAP). Conceptually the Fieldbus specification allows for the development of other Application Processes in the future, for example a PLC Application Process might be defined. In one field device, there are hundreds of parameters, such as the name of apparatus, an address, status variables and operating modes, function blocks, and those composed of data files. These parameters are defined as the objects in a VFD. They can be treated systematically, and are independent of the specification of the physical device. Each VFD is an "object" and within it there are other objects. An index of these objects, referred to as the Object Dictionary (OD) is provided within the VFD. It details each object within the VFD, their data types and definition. When another device, say a HMI host, wishes to access this data it can interrogate the VFD to determine what is available, its format etc.. This facility aids interoperability as well as automated configurability.

Point to Ponder: 5
A. What is the advantage of implementing a virtual field device? B. Why are most of the communications unidirectional in Fieldbus?

8. Fieldbus Devices
Field devices are control devices connected to Fieldbus network. They execute analog and discrete I/O functions plus the algorithms necessary for closed loop distributed control. From a communications perspective, field devices are composed of three components, namely, the function block application, the system management agent, and the communication stack, which includes the network management agent. This architecture for a field device, and its components, is described in detail by the Fieldbus Foundation Specifications. An overview is presented below.

8.1 Communications Stack
The communication stack of a field device is a three layer stack comprised of the Fieldbus physical, data link, and application layer protocols described above. The communication stack also contains a network management agent that provides for the configuration and management of the stack.

8.2 Transducer Block
Transducer Blocks may be output, input or a combination of the two. They interface between Fieldbus and the real world of sensors and actuators. An input Transducer Block converts signals coming from the plant into Fieldbus compatible variable and status messages. Output Transducer Blocks do the reverse. The content of Transducer blocks implementations are specific to the hardware technology they represent and consequently varies from vendors to vendors. They insulate function blocks from these specifics, making it possible to define and implement technology independent function blocks. Thus, while standardization is achieved through

Version 2 EE IIT, Kharagpur 15

function blocks described below, technological innovations in terms of electronics or signal processing is not stifled.

8.3 Resource Block
Resource Block contains the resource information for hardware and software within the Fieldbus device. For example, the device type, the manufacturer’s name and data such as, serial number and available memory capacity - are stored as parameters. Only one resource block exists in each Fieldbus device.

8.4 Function Block
The primary purpose of a field device is to perform low level I/O and control operations. The Function Block Application Process (FBAP), as defined by the Fieldbus Foundation, models these operations of a field device. The structure of an FBAP is shown in Figure 38.8. The FBAP is composed of a set of function blocks configured to communicate with each other. Outputs from one function block are linked to the inputs of another through configuration parameters called link objects. Function blocks may be linked within a device, or across the network. Function blocks are scheduled to execute their algorithms at predefined times that are coordinated with the transfer of their inputs and outputs. During the execution of the function block, the algorithm may detect events and trend parameter values (collect a series of values for subsequent reporting). Reporting of events and trends can be performed by multicasting them onto the bus to a group of devices. In addition some other types of objects such as View Objects, Alert Objects etc. may also be associated with an FBAP. These objects perform typical tasks related to the Function Blocks in the device. For example, • • • The 'Alert Object' monitors the status of various kinds of blocks, and reports to the upper system with a time stamp, if a configured alarm or event is detected. The ‘Trend Object' stores trend data within a device and sends it in one file upon request. This improves the communication efficiency. Similarly, the 'View Objects' construct dynamic files of variables, status indications etc., collected from various blocks, and required by external devices for monitoring and control purposes. By bundling together the required information it can be sent in a single transmission, thereby saving communication time. Finally, the 'Link Object'. Interfaces the function blocks and the objects to the FMS for implementing the configured Virtual Communication Channels between FBAPs residing within network devices

A function block is essentially a program that contains a set of network visible input parameters, output parameters, internally contained parameters, and an algorithm to process them. Parameters are identified by an index or a name (not recommended) that locates them in the object dictionary associated with the function block application. The object dictionary contains information used to encode and decode parameters, such as type and length, and also is used to map the parameter index to a local memory address. To promote interoperability, interface devices can access the object dictionary. Function blocks are connected to the physical hardware they represent through transducer blocks. Devices can be configured across the network through the use of contained block Version 2 EE IIT, Kharagpur 16

parameters. Contained block parameters are those that can be written to the device by interface devices. Interface devices are not able to write values to input and output parameters. Figure 38.10 shows how a communication relating to a physical variable takes place over the network. The value of the local physical variable is acquired by the function block through the transducer block firmware. This is processed by the function block and the output is communicated to another field device with the help of the Link Object. The Link object locates, from the object directory, the network address of the destination device as well as the mode of communication service to be used for the communication task. These are then realized by the FMS and the FAS sub-layers, in turn using the lower layers.
Input/Output Hardware Resources

Function Block Application Process Transducer Block Function Block View Objects Trend Object Link Object FMS Interface Link Object Alert Objects Link Object Resource Block

Clock Scheduling

Acyclic Communication Cyclic Communication

System Management Fieldbus

Field Device

Fig. 38.9 Architecture of a Function Block Application Process

Point to Ponder: 6
A. What is the difference between trend data and control data? B. Explain the i/o connections for alert, link, trend and view objects.

Version 2 EE IIT, Kharagpur 17

8.4.1 Realisation of Distributed Control Functions using Function Blocks in Fieldbus
Control functionality is realized over the network by a configured sequence execution of function blocks and communication tasks among them. For example, consider the control loop in Figure 38.6. The execution sequence is shown in Fig. 11. Note that there are three function blocks involved, namely, AI 101, PID 101 and AO 101. The first FB that executes is AI 101. This is followed by a cyclic communication of the process variable value to the PID 101 function block. The computation of the PID law in the FB PID 101 is followed by the computation of the valve stem position command to the positioner in the FB AO 101. Note there is no communication involved between these two FB executions, since both PID 101 and AO 101 are shown to be residing on the same Fieldbus device. Finally there is a communication between AO 101 and the host for the HMI station. This basic execution cycle repeats.
To other Function Blocks Input Parameter Physical Sensor Inputs Transducer Block

Output Parameters

Contained Parameters Object Dictionary Encoding/Decoding Info Object Name (opt) Object Index

Linkage Object

Read or Write Service: (Object Index, Request or Response Value)

Information Report Service: (Object Index, Value)

Fig. 38.10 Access to function block parameters through the object dictionary

Version 2 EE IIT, Kharagpur 18

Execution Cycle of PID Control Loop

AI101 - > PID101

AI101 - > PID101

AO101 - > Host

AO101 - > Host

AI 101 PID101 AO 101

Comm. with Host

Comm. with Host

Input Sampling

Valve Output Acyclic Communication

Function Block Execution

Cyclic Communication (Publish/Subscribe)

Fig. 38.11 Function Block execution and communication sequence for a control loop.

Point to Ponder: 7
A. Is there any impact of the delay between input sampling and valve output, as shown in Fig. 11? B. What is the relationship between the object dictionary and a function block?

9. Network Management Structure
Network management is the function of managing various parameters to carry out Fieldbus communication. Generally, execution of the communication function is performed by communications software that resides in a communications ASIC. The parameters for determining actual operation are called the Network Management Information Base (NMIB), and are grouped as one object. These parameters are accessed through Link Management Entity by the execution software at each layer. This function is transparent to the End User of the system.

10. System Management Functions
System Management (SM) performs the management of the parameters needed for the construction of a functional control system, rather than communication. The System Management Kernel is also modeled as an FBAP. The System Management Kernel performs two primary functions. The first is to assign End User defined names, called tags, and data link layer addresses to devices as they are added to the fieldbus. It contains an object dictionary and can be configured and interrogated using FMS operating over client/server virtual communication relationships.

Version 2 EE IIT, Kharagpur 19

The second is to maintain distributed application time so that function block execution can be synchronized among devices. Fieldbus has a common clock for called Link Schedule Time (LSTime). The LAS uses this to synchronize all devices on the bus frequently. Using LS-Time as a reference system management FBAP triggers each function block and synchronizes operation among Function Blocks in differing devices on the same bus. Furthermore, system management provides a real time reference, called Application time (AP-time). This time is used as the source of alarm or event time-stamps. To support these functions, the System Management Kernel communicates directly with the data link layer.

11. Device Description
Electronic Device Descriptions (EDDs) created by Device Description Language (EDDL) for a field device support the management of intelligent field devices. Typical tasks such as operation, parameterizing and diagnostics can thus be solved efficiently. EDD describes product features which serve as a basis for the entire electronic product data management from planning to engineering, set-up, maintenance and diagnostics and the disintegration of a plant. EDDs are ASCII files. They primarily contain the description of all device parameters and their attributes (e.g. lower/upper value range, default value, write rights) and device functions, e.g. for the plausibility check, scaling, mode changes or tank characteristics. EDDs also include a grouping of device parameters and functions for visualization and a description of transferable data records. Electronic Device Description Language (EDDL) is the mechanism that allows vendors to describe their products in a way that may be interpreted by any compliant host system. Thereby enabling compatibility and interoperability of devices. Also, the language allows vendors to include their specific product features while remaining compatible. Furthermore, the use of EDDL allows the development of new devices while still maintaining compatibility. The Device Description (DD) may be supplied with the device on a disk, or down loaded from the Fieldbus Foundation web site, and loaded into the host system.

Point to Ponder: 8
A. Device description languages enhance configurability and interoperability - justify or contradict. B. What is the difference between Link Schedule Time and Application Time?

Lesson Summary
In this lesson we have dealt with the following topics: A. Basic motivations for a networked communication architecture for distributed process automation: It is seen that several advantages can be realised, such as, reduced cabling, reliable and high speed communication, configurability, interoperability, enhanced diagnostics etc.

Version 2 EE IIT, Kharagpur 20

B. The communication architecture and its various layers: The three layers of the Fieldbus protocol have been discussed. The possible network topologies and media are described. It is explained how the requirements of real-time communication are met using a shared medium in the protocol. C. Architecture of Fieldbus devices: The software architecture of field bus devices is described. It is explained how the protocol is implemented within the device. The concept of Function and Transducer blocks are explained. The communication sequence involved in the distributed execution sequence of function blocks for realization of control functionality is explained. The mechanism of configuration using device description languages is introduced.

Version 2 EE IIT, Kharagpur 21

Answers, Remarks and Hints to Points to Ponder Point to Ponder: 1
A. How exactly does a network lead to reduced cabling over a point-to-point communication topology? Ans: The answer would be clear if you consider the sum of the lengths of line segments joining points on the circumference of a circle. This length grows proportional to the number of points. For a number of points more than 6, it crosses the length of the circumference of the circle. This is why the cabling requirement for point-to-point systems grows beyond that of a network. B. Can you cite three reasons why the reduction in cabling is considered a significant advantage of the Fieldbus? Ans: 1. Cost of data quality cables 2. Probability of interference and noise increasing with length of cable. 3. Complexity of maintaining a complicated hardware network of cables C. Can you state two advantages of interoperability? Ans: 1. Lower cost due to increased market competition 2. Better product functionality and quality of products D. For what kind of factories is a Fieldbus implementation for plant automation justified? Ans: For large factories employing sophisticated manufacturing processes, where the return on investment can be justified.

Point to Ponder: 2
A. Why are network and transport layers absent in a Fieldbus network? Ans: Because the routing of messages in the network is fixed. Therefore the network layer is redundant. Also since the traffic is nearly fixed, transport layer functionality, such as flow control, are redundant. B. Why are session and presentation layers absent in a Fieldbus network? Ans: Because the task composition in the system and their communication requirements are fixed at configuration time. Also because the limited requirements of the layer are clubbed into the Application Layer.

Version 2 EE IIT, Kharagpur 22

Point to Ponder: 3
A. Why are repeaters required for extending the bus in a network ? Ans: Repeaters clean up the distortion of digital data which increases with the length of the transmission channel. B. Why does the maximum bus length depend on the transmission rate ? Ans: Firstly because of increasing series mode interference along the channel. Secondly because of the increasing load capacitance seen by the channel drivers which needs to be driven at the higher data rates.

Point to Ponder: 4
A. How does the LAS know how much time to give to which process and in what sequence? Ans: This is prefixed at configuration time by the configuration of the function blocks in the system. B. Is the proposed medium access strategy adequate for industrial automation? Ans: Yes it is. Since the sampling and control update rates are sufficiently low. C. Can you think of an alternative strategy for network communication? Ans: There are several other real-time networking protocols employed by networks such as Devicenet, Canbus etc.

Point to Ponder: 5
A. What is the advantage of implementing a virtual field device? Ans: The advantage is that the network device tags and other communication parameters such as addresses, communication modes etc. can be made transparent to the high level function blocks. B. Why are most of the communications unidirectional in Fieldbus? Ans: Firstly because the nature of the typical application tasks are unidirectional. For example a transmitter only transmits data out to controllers. Secondly, communication takes place over short network channels and therefore in most cases data is received properly and the requirements for acknowledgements necessitating bidirectional communication does not exist.

Version 2 EE IIT, Kharagpur 23

Point to Ponder: 6
A. What is the difference between trend data and control data? Ans: Trend data is generally averaged over several control data samples. Also requires a slower update rate. B. Explain the i/o connections for alert, link, trend and view objects. Ans: Alert objects report malfunctions in the device itself or abnormalities in the process. Thus it requires input from the hardware resources and function blocks. Link objects take input from all other blocks and outputs these in appropriate forms to the FMS. Trend and objects compute trend and view data based on the control data. Thus they take input from the function blocks.

Point to Ponder: 7
A. Is there any impact of the delay between input sampling and valve output, as shown in Fig. 11? Ans: Introduces a time delay into the loop. However, the value of this delay is generally negligible in comparison with the dominant time constants of the control loop. B. What is the relationship between the object dictionary and a function block? Ans: The detailed communication and other parameters corresponding to function block parameters is stored in the object dictionary.

Point to Ponder: 8
A. Device description languages enhance maintainability - justify or contradict. configurability, interoperability and

Ans: Yes it does. The Device description is read by a host to re-configure the system to include the device and support its communications. Interoperability is enhanced, since the DDL provides a uniform syntax for describing products from all vendors. B. What is the difference between Link Schedule Time and Application Time? Ans: Link schedule time is relative time for scheduling tasks over elementary and macrocycles. Application time is running time until it is reset by specific events defined for the device.

Version 2 EE IIT, Kharagpur 24

Module 9
Conclusion
Version 2 EE IIT, Kharagpur 1

Lesson 39
Higher Levels of Automation Systems
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
After learning the lesson students should be able to A. Describe the major functions of Production Management Systems under Level 3 Automation B. Describe the major features of a Supervisory Control System under Level 2 Automation C. Describe the major features of a Distributed Control System (DCS)

Introduction
The major activities that define the manufacturing enterprise in direct relation to the manufacturing of products include (i) Production Management, and (iii) Manufacturing. The first function is incorporated in Level 3 automation while the second one refers to Level 1 and Level 2 functions of a hierarchical Industrial Automation System. In the earlier lessons of this course we have predominantly been concerned with Level 1 Automation for manufacturing. In this lesson we briefly discuss these other aspects and particularly Level 2 and Level 3 Automation in more detail below. Finally a typical implementation of the functionality of Level 2 and 3 automation, under the technological platform known as Distributed Control Systems (DCS).

Level 3 Automation: Production Management
The role of production management is to plan and control effectively the physical and operational resources of the manufacturing plant such as, materials, tools, fixtures, machines, storage space, material handling equipment, and manpower, so as to meet the production requirements. Production management is also referred to as Production Planning and Control. As the name implies, it comprises two functions, namely, Production Planning and Production Control. Production planning is concerned with: (1) deciding on the set of products to be manufactured, along with their production volumes, over a certain duration which is called the planning horizon; (2) scheduling, i.e., determining the sequence of production of the set of parts and products over the time duration and (3) allocation of the necessary manpower, raw material and equipment resources needed to accomplish the production plan. Activities within the scope of production planning include the following. • Aggregate production planning. This involves determination of the target production output levels. The word aggregate means that planning is conducted at a gross level to meet the total demand collected over all products consolidated into product groups, utilizing the total human, equipment and material resources. These plans must be made in cognizance of various functions of the firm other than manufacturing, such as product design, production, inventory, marketing, and sales for the plan to be feasible and effective for the overall business objectives of a firm. Further, it must be made based on an accurate forecast of the demand for the products. Master production planning. In the next stage, the aggregate production plan must be detailed into a master production schedule (MPS), which includes specification of target production volumes of individual product types and their production schedules. In turn, Version 2 EE IIT, Kharagpur 3

this master schedule must be converted into purchase orders for raw materials, orders for subcontracting, and product schedules for subassemblies and components. These activities must sequence properly and coordinated to enable the delivery of the final product on schedule, under the existing resource constraints. Based on the MPS, one now carries out the allocation of production resources in the following two steps. Typical horizons for MPS may be of several months. • Material requirements planning (MRP) translates the MPS of end products into a detailed schedule for the procurement of raw materials and parts used in those end products. More precisely, MRP takes the master schedule data, bill of materials file, and the inventory data and determines when to order raw materials and components for assembled products. It can also reschedule orders in response to variations in production priorities and demand conditions. Capacity planning is concerned with determining the labor and equipment resources needed to achieve the master schedule. Capacity planning can be carried out, either on a short-term or a long-term basis. Short-term capacity planning decisions include: overtime or reducing the workweek, hiring and firing, subcontracting, and inventory stockpiling. Long-term capacity planning includes decisions such as acquisition of new resources and manpower, augmentation or closure of facilities etc. Typically capacity and materials plans are made for periods of weeks to a month.

Production control is concerned with providing for the necessary resources to implement the production plan. The major aspects of production planning include: • Shop floor control transforms the planning decisions into control commands for the production process. It also involves collection of data related to shop floor operations and processing and communication of the data on to higher control levels. Also termed as, Production Activity Control, it consists of three activities: order release, order scheduling, and order progress. o The purpose of the order release module is to provide the necessary documentation that accompanies an order as it is processed through the shop floor. This documentation consists of route sheet, material requisitions, job cards, parts list, etc. o Order scheduling involves assignments of orders to various machines and manufacturing units, so that the delivery schedules are met, in-process inventory is minimized, and machine utilization is maximized. Order scheduling involves two steps: (1) machine loading, and (2) job sequencing. Machine loading involves allocating jobs to machines, and job sequencing determines the order in which the jobs are processed through a work center. Several priority rules could be used for job sequencing. These include shortest processing time, first-come-first-served, least slack, etc. o The purpose of the order progress module is to continuously acquire and communicate data relating to work-in-process and shop order status. • Inventory control. Inventory control is concerned with maintaining certain levels of stock for raw material, semi-finished parts, subassemblies, and finished goods. This is needed to create a buffer between the company and its suppliers and consumers and also between different stages of the manufacturing system. Such a buffer can provide insurance against Version 2 EE IIT, Kharagpur 4

machine failures, uncertain demands, uncertain suppliers, and worker absenteeism so that manufacturing activities do not get stalled. However, the inventory should be kept at an optimal low at the same time maintaining satisfactory customer service. The activities in a modem PPC system and their interrelationships are depicted in Figure 39.1. A sophisticated automation system that combines MRP and capacity planning as well as shop floor control and other functions related to PPC is known as Manufacturing resource planning and a standard for the same is called MRP II. As the figure indicates, PPC ultimately extends to the company’s supplier base and customer base. This expanded scope of PPC is known as supply chain management.

Points to Ponder: 1
A. How is ERP (short for Enterprise Resource Planning), popular software based technology related to Level 3 automation ? B. The above description of Level 3 automation is given mainly in the context of discrete manufacturing. Do you expect different features for continuous manufacturing processes, such as a refinery or a steel plant ?

Level 2 Automation: Supervisory Control
Supervisory control combines the firm’s production scheduling and management information functions with the process control functions to form a hierarchical control system. Figure 39.2 outlines a typical functional hierarchy of such an industrial computer control system. It should be noted that the several levels shown in Figure 39.2 are operational levels and do not necessarily represent separate and distinct computational hardware levels. In large systems a separate computer may be needed to handle each level, but in small systems, two or more operational levels might be collapsed into one computer level. The dedicated digital controllers at Level 1 require no human intervention since their functional tasks are completely fixed by systems design and are these are not interacted with, on-line, by operators. All other levels have human interfaces as indicated.

Version 2 EE IIT, Kharagpur 5

Sales and marketing

Aggregate production planning Aggregate planning Master production schedule Inventory Detailed planning Material requirements planning Capacity planning

Product design

Engineering data base

Production control

Purchasing

Shop floor control

Operations

Supplier

Factory

Customer

Fig. 39.1 Activities for Production Planning and Control and their relationships with other functions in the firm, customers and suppliers

Version 2 EE IIT, Kharagpur 6

(level 3)

Management information visualisation

Production Planning and Control Data

Communications with other supervisory systems

(level 2)

Supervisor’s console

Supervisory Control Data

Communications with other DDC systems

(level 1)

Operator’s console

Direct Digital Control Data Plant inputs

(level 1) Plant inputs

Stand-alone Process Controllers Control Output Process

Fig. 39.2 Functional hierarchy of an industrial control system For a large manufacturing complex there is need not only for vertical command/feedback communication among the levels, there is also need for horizontal level communications for coordination among automation systems for various units. Such communications are shown in Fig. 39.3. The level 2 and level 3 automation functionalities are therefore split into two distinct blocks, namely, 2A/2B and 3A/3B. While the blocks 2A and 3A interact vertically within the same area control system, the blocks 2B and 3B interact horizontally across operational areas.

Version 2 EE IIT, Kharagpur 7

(level 3B)

Management Information visualisation

Production Planning and Control

Communications with other areas

(level 3A)

Operational and production

Production scheduling and operational management

Communications with Supervisory System

(level 2B)

Supervisor’s console

Intra-area coordination

Communications with Supervisory Systems of other areas Communications with other DDC systems

(level 2A)

Supervisory Control

(level 1)

Operator’s console

Direct Digital Control

(level 1)

Stand-alone digital controllers

Process Fig. 39.3 Functional hierarchy of a multi-area industrial control system The Level 2 automation systems offer the following two main capabilities: 1. Tight optimized control of each operating unit of the plant based upon the production levels and constraints set by Level 3 PPC system by providing optimal operating set points to the manufacturing processes. This control reacts directly to any emergencies that occur in its own unit. 2. Improved overall reliability and availability of the total control system through fault detection, fault tolerance, redundancy, and other applicable techniques built into the system’s specification and operation.

Version 2 EE IIT, Kharagpur 8

Supervisory Control Tasks
In the context of a large industrial plant, the tasks carried out at each level of the hierarchy are as described in Table 39.1. A. Control Enforcement 1. Respond to any emergency condition in its scope of plant operation. 2. Locally optimize the operation of units under its control within limits of established production schedule; Plant coordination and Information management 3. Collect and maintain data related to production, inventory, and raw material and energy usage for the units under its control. 4. Maintain communications with higher and lower levels. 5. Service man/machine interfaces System Reliability Assurance 6. Perform diagnostics on itself and lower-level machines. 7. Manage system redundancies. Table 39.1 Main tasks of the Supervisory Level Details of how the operations are actually carried out will vary significantly, particularly at the lowest levels, because of the nature of the actual processes being controlled. To perform the above tasks, the system must sample each important plant or process variable continuously. Plant variables are normally sampled at variable rates depending upon the type of variable being sensed, and the dynamics or speed of response of the process being and monitored and controlled. The sampled values of the variables are processed to compute diagnostic features, such as average/rms values and compared with a set of thresholds to detect the presence of any abnormal operating condition. These may also be compared with operating set point to compute controls. In addition to being used for emergency detection and control corrections, the values of the process variables are stored in the computer’s memory in a process database. These data are then used for operator’s console read-out functions, for data logging for historical records, for process efficiency calculations, and for read-out to higher level computers for optimization calculations, inventory monitoring, overall plant production status and historical file updates, and other necessary calculations. An equally important task is to maintain the plant’s production data base for the company’s production, financial, and personnel reports.

B.

C.

Points to Ponder: 2
A. Note that while Level 1 automation is mostly free of human intervention, in Levels 2 and 3, appreciable human interaction with automation systems may be needed. Considering this what do you think would be some desirable features of man-machine interfaces for these levels ?

Version 2 EE IIT, Kharagpur 9

B. “It can be said that Level 1 and Level 3 automation technology largely depends on the core technological areas of Process Data Analysis and Optimization.” Provide your comments on the above statement based upon studies of industrial applications Finally we discuss the technology platform that realizes the Level 1 and Level 2 automation, namely Distributed Control Systems (DCS)

Distributed Control Systems (DCS) Brief History
Distributed control system (DCSs) have been evolving rapidly since the mid 1980s from being essentially panelboard replacement at their inception to become comprehensive plant information, computing, and control networks fully integrated into the mainstream of plant operations. This progress has been fueled in part by the technological revolution in microprocessor and software technology and in part, by economic necessity. Microprocessor-based DCSs made their debut in the mid-1970s. Initially they were conceived as functional replacements for electronic panelboard instrumentation and were packaged accordingly. The initial systems utilized discrete panelboard displays similar to their electronic instrumentation counterparts. These systems evolved quickly, adding video-based workstations and shared controllers capable of expressing complex unit-operations-oriented regulatory and sequence control strategies containing scores of functional elements, such as PID (proportionalintegral-derivative), lad/lag/totalizers, dead-time elements, elapsed timer, logic circuits and general-purpose calculators By the early to mid-1980s the personal computer industry matured into a multibillion dollar per year marked with the IBM PC disk operating system (DOS) as the standard. This gave birth to the software industry that delivered feature-laden high quality inexpensive software packages. The opportunity for system integrators and value-added resellers was clear. One could devolop an relatively-inexpensive scan control alarm, and data acquisition (SCADA) package for a personal computer platform and integrate it with these general-purpose shrink wrap software packages, such as spreadsheet, desktop publishing, or database management, and one could have a very cost-effective alternative to DCS. Because of performance and the general suitability limitations of these PC offerings, this approach had appeal mostly in cost-sensitive noncritical applications and where these existed a low safety or hazard risk. This concept, however, created an expectation and vision of the future, that is, open architectures. DCS vendors felt a compulsion to enrich their arsenal of tools to address real-time process control applications by incorporating the low-cost shrink-wrap packages into their systems. Such packages included: ♦ Relational database management ♦ Spreadsheet packages ♦ Statistical process control capabilities ♦ Expert systems ♦ Computer-based process simulation ♦ Computer-aided design and drafting

Version 2 EE IIT, Kharagpur 10

♦ Desktop publishing ♦ Object-orineted display management ♦ Windows-oriented display management ♦ Information exchange with other plant systems. During the last 1980s and early 1990s the computer industry continued its transformation. Networking of systems into a cohesive whole promised to (again) revolutionize an industry, which has barely absorbed the impact of the PC revolution. Software and communications standard began to take hold, making interoperability among disparate computing platforms and application software a near-term reality. The business enterprise, including the factory floor, could be molded into a cohesive whole by making the various departmental systems work cooperatively at an acceptable integration cost. These added new technological features to DCS including: ♦ Open operating system standards, such as UNIXC ir POSIX ♦ Open system interconnect (OSI) communiations model ♦ Client server cooperative computing model ♦ X-window protocols for workstation communications ♦ Distributed relational database management systems ♦ SQL access to distributed relational databases ♦ Object oriented programming and platform independent languages ♦ Computer-aided software engineering (CASE) These characterize the modern DCS technology. DCSs, today are distributed computing platforms with sufficient performance to support large-scale real-time process applications. Structurally DCSs traditionally are organized into five major subsystems, namely (1) operations workstations that act as the MMI and provide visualisation capability, (2) controller subsystems that perform direct digital control, (3) data collection subsystems, (4) process computing subsystems for process optimisation and supervision, and (5) communication networks. Open system communication standards are enabling DCSs to receive information from a set of similar compatible computing platforms, including business, laboratory information, maintenance, and other plant systems as well as to provide informations in support of applications, such as: ♦ Automated warehousing and packaging line systems so that a complete order can be coordinated from the receipt of raw materials to the shipment of the final product. Laboratory information management systems (LIMs), which perform in-process analysis as well as quality assurance inspections. ♦ Automated production scheduling for a plant accessing the business system and tying into MRP II systems and finite-capacity scheduing packages.

Version 2 EE IIT, Kharagpur 11

Appendix A: An Example Functional Specification document for Basic Level (Level 1) and Process Control Level (Level 2) Automation Systems for a large rolling mill
Platforms: The above levels of controls shall be achieved through programmable controllers PLCs, micro-processor based systems as well as PCs / Work stations, as required. Each of the automation systems of the Plant shall be subdivided in accordance with the functional requirements and shall cover the open loop and closed loop control functions of the different sections of the line and the mill. Modes of Operation: The systems shall basically have two modes of operation. In the semiautomatic mode the set point values shall be entered manually for different sections of the line through VDU and the processors shall transmit these values to the controls in proper time sequence. In fully automatic mode the process control system shall calculate all set point values through mathematical models and transfer the same to the subordinate systems over data link. Functionality at Basic Level (Level 1): The Basic Level shall cover control of all equipment, sequencing, interlocking micro-tracking of strip for specific functions, dedicated technological functions, storage of rolling schedules and look-up tables, fault and event logging etc. Some of these are mentioned below. ♦ All interlocking and sequencing control of the machinery such as for entry and exit handling of strips, shear control etc. Interlocking, sequencing, switching controls of the machines. This shall also cover automatic coil handling at the entry and exit sides, automatic sequencial operation of welding machine and strip threading sequence control as well as for acid regeneration plant. ♦ Calculation of coil diameter and width at the entry pay-off reels. ♦ Position control of coil ears for centrally placing of coils on the mandrels. ♦ Generation of master speed references for the line depending on operator's input and line conditions and down loading to drive control systems. ♦ Speed synchronising control of the drives, as required. ♦ Strip tenstion, position and catenary control through control of related drives and machinery. ♦ Initiation of centre position control for PORs, steering/dancer rolls; Looper car position control. Automatic pre-setting control, measurement and control of tension and elongation for tension leveller. Auto edge position control at tension reels if required. ♦ Control of entry shear for auto-cutting of off-gauge strip. ♦ Control of pickling parameters for correct pickling with varying speed of strip in the pickling section. ♦ Side trimmer automatic setting contro. ♦ Interlockings, sequencing and control of scrap baller, if provided. ♦ Auto calibration for position control/precision positioning shall be provided as necessary. Version 2 EE IIT, Kharagpur 12

♦ Manual/Auto slowdown/stoppage of strip at weld point at tension leveller, side trimmer, mill and exit shear. ♦ Control of technological functions for tandem mill such as : o Automatic gauge control along with interst and tension control. o Shape control o Roll force control ♦ Storage of tandem mill rolling schedules, for the entire product mix and all possible variations. Suitable look-up tables as operators guidance for line/equipment setting. ♦ Automatic roll changing along with automatic spindle positioning. ♦ Constant pass line control based on roll wear as well as after roll change. ♦ Automatic control of rotary shear before tension rells. ♦ Automatic sequence control of inspection reel. ♦ Provision of manual slow down/stoppage of strip as well as chearing for `run' for inspection of defects at tension leveller, side trimmer entry and exit of the Tandem Mill throuth push button stations. ♦ Micro-tracking of strip and flying gauge change (set point change) for continuous operation with varying strip sizes. ♦ Setting up the mill either from the stored rollings schedule with facility for modification by the operator of down-loading from process control level system. ♦ Automatic control of in-line coil weighing, marking and circumferential banding after delivery tension reels. Supervisory Functions at Basic Level A: Centralised supervisory and monitoring control system shall be provided under basic level automation with dedicated processors and MMI. All necessary signals shall be acquired through drive control system as well as directly from the sensors/instruments as, required. The system shall be capable of carrying out the following fuctions. ♦ Centralised switching and start up of various line drives and auxiliary systems through mimic displays. ♦ Status of plant drives and electrical equipment for displaying maintenance information. ♦ Monitoring and display of measured values for tandem mill main drives and other large capacity drives such as winding temperature, for alarm and trip conditions. ♦ Centralised switching and status indication of 33 kV and 6.6 kV switchboards. ♦ Display of single line diagram of 33 kV and 6.6 kV switchboards, main drives, in-line auxiliary drives etc. ♦ Acquisition of fault signals from various sections of the plant with facility for display and print-out of the fault messages in clear text.

Version 2 EE IIT, Kharagpur 13

Functionality at Process Control Level: The Process Control Level shall be responsible for computation and control for optimization of operation. Functions like set point generation using mathematical models, learning control, material tracking within the process line/unit including primary data input, real time control of process functions through basic level automation, generation of reports etc. shall be implemented through this level of automation. Some of the specific functions to be performed by the process control level automation are the following. ♦ Coil strip tracking inside the process line/unit by sensing punched holes at weld seams. ♦ Primary Data Input (PDI) of coils at entry to PL-TCM with provision for down loading of data from production control level. ♦ Generation of all operating set points for the mill using PDI data, mill model, roll force model, power model, strip thickness control model, shape/profile control model with thermal strip flatness control as well as for other sections of the line. ♦ Learning (Adaptive) control using actual data and the mathematical model for set-up calculations. ♦ Storage of position setting values of levellers, side trimmer. Input of strip flaw data manually through inspection panel at the inline inspection facility after side trimmer. ♦ Processing of actual data on rolling operation, generation of reports logs and sending data to production control level. Information System Functions: The information system shall generally comply with the following features. ♦ Data of importance shall be available with the concerned personnel in the form of logs and reports. ♦ Output of logs and reports at preset times or on occurance of certain events. ♦ It shall be possible to change the data items and log formats without undue interference to the system. ♦ Logged information shall be stored for adequate time period ensuring the availability of historical data record. ♦ Data captured by the system shall be checked for integrity with respect to their validity and plausibility with annunciation. Man Machine Interface: The visualisation system for both the automation levels shall be through man-machine interface (MMI) for the control and operation of the complete line. The system shall display the following screens, with facilities for hard copy print out. ♦ Process mimics for the complete line using various screens with status information of all important in-line drives as well as the references and actual values of important parameters. ♦ Dynamic informations in form of bar graph for indication of reference and actual values of important parameters. ♦ Screens providing trends of the important process variables. Version 2 EE IIT, Kharagpur 14

♦ Acquisition of actual parameters (averaging/maximum/minimum) for the complete line, on coil to coil basis through weld seam tracking or TCM exit shear cut for the generation of logs on process/parameters and production. Standards: The programmable controllers and other micro processor based equipment offered shall generally be designed/structured, manufactured and tested in accordance with the guidelines laid down in IEC-1131 (Part 2) apart from the industry standards being adopted by the respective manufactures. Hardware: The hardware of each basic controller/equipment of a system will generally comprise main processing unit, memory units, stabilised power supply unit, necessary communication interface modules, auxiliary storage where required. I/O modules in the main equipment, remote I/O stations where required and the programming and debugging tool (PADT). The hardware and software structure shall be modular to meet wide range of technological requirements. I/Os shall be freely configurable depending on the requirement. The programming units shall preferable be lap-top type. Networking: The networking would conform to the following specifications. ♦ In each of the two automation levels, all the controllers of a system shall be connected as a node over suitable data bus forming a LAN system using standardised hardware and software. ♦ The LAN system shall be in line with ISO-Open system Interconnect. ♦ All drive level automation equipment shall be suitably linked with the basic level for effective data/signal exchange between the two levels. However, all the emergency and safety signals shall be directly hardwired to the respective controllers. ♦ Similarly, the LAN systems for the basic level and process control level shall be suitably linked through suitable bridge/interface for effective data/signal exchange. Provision shall also be made for interfacing suitably the process control level with the production level automation system specified in item. ♦ The data highways shall be designed to be optimally loaded and the same shall be clearly indicated in the offer. ♦ The remote I/Os, the microprocessor based measuring instruments and the microprocessor based special machines like coil weighing, marking and circumferantial banding machines shall be connected over serial links with the respective controllers. ♦ The personal computers and work stations shall be connected as a LAN system of the corresponding level. Data and Visualisation: The following specifications would apply in respect of data security, validity and its proper visualisation. ♦ All the operator interfaces comprising colour VDU and keyboard as MMI for interacting with the respective system and located at strategic locations, shall be connected to the corresponding LAN system. ♦ Keylock/password shall be provided to prevent unauthorised entry. ♦ Entry validity and plansibility check shall also be incorporated. Version 2 EE IIT, Kharagpur 15

♦ An Engineer's console comprising of necessary processor, color VDU, keyboard/mouse and a printer unit shall be provided for the automation systems. The console shall have necessary hardware and software of communicating with the LAN and shall have access to the complete system. Basic functions of this console shall be off-line data base configuration, programme development, documentation etc. Application Software: The application software shall be through functional block type software modules as well as high level language based software modules. The software shall be user friendly and provided with help functions etc. Only one type of programming language shall be used for the complete system. However, ladder type programming language may be used for simple logical functions. Only industrially debugged and tested software shall be provided.

System Figures of Merit
Future Expandibility: The selection of equipment, standard software and networking shall be such as to offer optimum flexibility for future expansion without affecting the system reliability. Fault Tolerance: The system shall be designed to operate in automatic or semi automatic mode of operation under failure conditions. Spare Capacity: The system shall have sufficient capacity to perform all functions as required. A minimum of 30 per cent of the total memory shall be kept unallocated for future use. Loading: The data highway shall be designed to be optimally loaded and the same shall be clearly indicated in the offer. Software Structure and Quality Programs: shall be in high level language that is effective and economical for the proposed system in respect of Modularisation, rate of coding, store usage and running time. The software structure of the system shall be suitably distributed/centralised for supervision and control of the related process areas following the state of the art architecture. Integration: The communication software shall be such that the systems shall be able to communicate independently among themselves as well as with the lower level Basic Control/Process control automation system, as required. Provision shall be made for interfacing the production control system with the higher level Business Computer system to be provided for the entire steel plant in future. Programmability: The information system shall generally be designed such that it shall be possible to change the data items and log formats without undue interference to the system. Data Integrity and Protection: Logged information shall be stored for adequate time period ensuring the availability of historical data record. Data captured by the system shall be checked for integrity with respect to their validity and plausibility with annunciation. Storing of essential data to be protected against corruption when the system loses power supply or during failure.

Points to Ponder: 2
A. The above specification document was developed while ordering for the automation system solution for an advanced rolling mill. Compare the functional divisions of the

Version 2 EE IIT, Kharagpur 16

Basic Level and the Process Levels with that of Levels 1, 2 and 3 as discussed in this course. B. Name three additional functionality requirements of automation systems that you may not have found earlier.

Appendix B: Features of an Industrial DCS: Honeywell’s' Total Plant Solution (TPS) System
Honeywell offered a new generation DCS system christened Total Plant Solutions (TPS) in 1996. The fundamental characterizing feature of the system is the unified environment for business and control. The system is highly integrated and Honeywell has provided solution from sensing to real-time control to process optimization to batch scheduling and inventory control. In other words the same system supports all level 0 - level 3 function without any very distinct structural boundary. The integration is achieved through a virtual network, which seamlessly integrated a PC based plant wide intranet such as Ethernet, and the process networks such as Honeywell’s LCN and UCN or Foundation Fieldbus. Technologically, it uses state of art hardware and software standards such as window NT, OLE, Fieldbus, Power PC/Intel chips, ODBC (Open Database Connectivity). The new components in the TPS system are the following. ♦ Global User Station ♦ Total Plant Desk Top ♦ Total plant equipment tools ♦ Total plant history ♦ Advanced suite of application software for Multivariable Control ♦ Optimization and Batch process Management Global User Station (GUS) is the interface to the open business environment. (GUS) is designed to be the Windows NT GUI for all Honeywell Control Systems - TPS as well as TDC 3000. GUS is a Power PC and a process network processor (LCN/UCN). Power PC run all window NT software. The network processor, which is a PCI bus plug-in-board on the Power PC, controls the secure control environment. It runs Honeywell’s RNOS (Real-time Network OS). Thus it is possible to organize and operator interface controlling data both from the Business and the Control Environments. Application software is all based on window WT and support all Microsoft Tools such as OLE, SQL, OBDC, netDDE and video. TPS manages a common database, which shared by controllers in client server architecture. For realizing Level 1 functions TPS has a High Performance Process Manager, Fail Safe Controllers and supports high density I/O, the LCN and fieldbus networks. Between the open environment and the secure control environment is a `firewall’, which prevents the degradation of real-time control by other application due to excessive intrusion and resultant constants improved on resources such as network bandwidth or CPU time. The following are among the Software Packages provided. ♦ Historian ♦ Advanced Control Solution for a range of Industry Application. Version 2 EE IIT, Kharagpur 17

♦ Desktop tools provides tightly integrated interfaces to Microsoft ♦ Excel and Access for Power Trend Analysis. Honeywell has entered into a strategic alliance with SAP American Inc for integration of business modules for sales procurement, order entry, production invoicing and shipping, with TPS for a tight enough to business and control

Points to Ponder: 3
A. Name three areas of software technology that have significant impacts on modern DCS solutions. B. What are the advantages of Open Standards compared to Proprietary Standards, in the context of Automation systems ?

Lesson Summary
In this lesson, the following topics related to higher levels of automation have been discussed. A. Production Planning and Control B. Supervisory Control C. Distributed Control Systems (DCS) D. Example functional specifications for Level 2 and Level 3 automation of a large industrial process E. Features of a modern Industrial DCS system

Version 2 EE IIT, Kharagpur 18

Answers, Remarks and Hints to Points to Ponder Points to Ponder: 1
A. How is ERP (short for Enterprise Resource Planning), popular software based technology related to Level 3 automation? Ans: ERP is a software technology to integrate all aspects of information related to operation in an organization including its manufacturing and business aspects. It is thus the enabling technology on which Level 3 and Level 4 automation systems can be implemented and integrated. This is evident from the fact that modern DCS systems support interfaces to standard ERP products. B. The above description of Level 3 automation is given mainly in the context of discrete manufacturing. Do you expect different features for continuous manufacturing processes, such as a refinery or a steel plant? Ans: For continuous manufacturing, such as refinery and steel, generally, product variety is lower, while production volumes for are high for individual products. Variations in product levels are also slower. It is for this reason, that, longer range scheduling is possible and is somewhat easier. On the other hand, these processes being highly energy intensive and requiring continuous material transformation, synchronism between operations in various units can have big impact on quality and productivity. In general it can be said that, aspects of Level 3 automation may be simpler in process plants, but aspects of Level 2 automation tend to be more complex.

Points to Ponder: 2
A. The above specification document was developed while ordering for the automation system solution for an advanced rolling mill. Compare the functional divisions of the Basic Level and the Process Levels with that of Levels 1, 2 and 3 as discussed in this course. Ans: As mentioned in the specification document, the Basic Level shall cover control of all equipment, sequencing, interlocking micro-tracking of strip for specific functions, dedicated technological functions, storage of rolling schedules and look-up tables, fault and event logging etc. Centralised supervisory and monitoring control system shall be provided under basic level automation with dedicated processors and MMI. All necessary signals shall be acquired through drive control system as well as directly from the sensors/instruments as, required. The Process Control Level shall be responsible for computation and control for optimization of operation. Functions like set point generation using mathematical models, learning control, material tracking within the process line/unit including primary data input, real time control of process functions through basic level automation, generation of reports etc. shall be implemented through this level of automation. Therefore, basic level includes both Level 1 and Level 2 functions. Process Control Level includes Level3 functions and some functions that may be called Level 2. Version 2 EE IIT, Kharagpur 19

B. Name three additional requirements of automation systems that you may not have found earlier. Ans: Apart from functionality requirements, there are several IT related require on automation system products. Some of these mentioned above are: Future Expandibility, Software Quality, Data Integrity and Protection

Points to Ponder: 3
A. Name three areas of software technology that have significant impacts on modern DCS solutions. Ans: Database, Networking and Decision Support B. What are the advantages of Open Standards compared to Proprietary Standards, in the context of Automation systems? Ans: Open Standards promote competitiveness among vendors and provides maximum value to the customer, in terms of functionality, quality, reliability and cost of automation systems. Adoption of such standards imply that the customer can purchase equipment from a variety of vendors and integrate into the over all system without difficulty. Proprietary standards were followed earlier by big automation vendors and have largely been replaced by open standards today.

Version 2 EE IIT, Kharagpur 20

Module 9
Conclusion
Version 2 EE IIT, Kharagpur 1

Lesson 40
Conclusion and Review
Version 2 EE IIT, Kharagpur 2

Instructional Objectives
This is the concluding lesson of the course. It attempts at providing a brief summary of the lessons discussed. It also attempts to highlight the important features of different lessons

Assessment and Review of the Course
Before a programme of instruction ends, it is important, not only to review what have been discussed, but also to highlight some of the topics within the area of study that may have been left out. It is also important to discuss, how an interested learner can move ahead from where the present programme ends. It is the main purpose of this lesson to address the above issues. Industrial Automation is a highly interdisciplinary area, if one intends to go deep enough to the level of a system designer. It is because of this reason, that it is impossible to achieve this depth for any audience with an undergraduate background in any major engineering discipline, not to mention the limitations of the course author. Therefore, the course is restricted to a depth, where the aim is to describe how things work, along with some of their technical features related to system performance. It is not attempted to discuss issues related to design, manufacture and test of such equipment. Depth of exposition has also been sacrificed with respect to the hierarchy of Industrial Automation Systems. This course is primarily concerned with Level 0 and Level 1 Automation. The discussions on Level 2 Automation, which encompasses Process Monitoring, Supervision and Operational Optimisation, as well as Level 3 Automation which encompasses Process Operations Planning and Management have been very limited. Within the limitations in scope and depth, the course aims to cover most important topics of Level 0 and Level 1 Automation. It provides a reasonably complete treatment of major types of sensors of important process variables, such as temperature, pressure, flow, which are of importance in continuous processes, as well as mechanical variables such as position, velocity, acceleration, force, etc., which are of importance in discrete manufacturing industries. Similarly, hydraulic and pneumatic actuators, control valves, electric motors and drives, pumps and fans have been described to provide coverage on industrial actuation systems. One of the most interesting flavors that this course offers to the learner is a flavor of practice. Thus, while a conventional course on Control Systems provides a lot of analytical insight into the working of feedback controlled systems, these generally do not include interesting control structures such as the Smith Predictor which provide significantly superior performance for plants with practical features such as transportation lags. Similarly, they do not discuss practical implementation features to address real-world issues as actuator saturation or auto-manual transfer. Similarly, conventional courses on Process Control are generally solely concerned with continuous processes. Control problems related to discrete manufacturing are generally discrete in nature, and often involve significant extent of discrete sequence and logic control. Among continuous control functions, they generally require precise control of position, speed, force etc. Moreover, they do not treat issues related to actuation, in depth. Actuation systems however are often so complex and substantive that they deserve an independent treatment at some level of Version 2 EE IIT, Kharagpur 3

detail. Without an understanding of actuation, it is difficult to understand the functionalilty and performance of control systems. In this course actuation systems have been allocated adequate space to enable an appreciation of the working of automated industrial operations. However, the treatment is restricted to hierarchically lower levels of the actuation system, such as drives and hydraulics. Based on these one can appreciate high level industrial machines such as Robots, Automated Guided Vehicles, Cranes, Metal forming and Cutting machines. After basic courses on Control, Process Control and Instrumentation, this course is expected to provide the reader with a more complete and integrated understanding of industrial automation technology. Another important aspect of the course is the modular structure of the lessons. Broadly speaking, the different modules can be identified as follows: • Introduction, • • • • • • • Industrial Instrumentation, Process Control, Programmable Logic Control Systems, CNC Machines and their control, Valves and Actuators, Electrical Machine Drives, Industrial Embedded and Communication Systems.

The modular structure will possibly help the reader to skip some of the modules, if he is familiar with them. Moreover the contents have been written in a way suitable for self-study. Review questions and Points to ponder would help the reader to bring clarity in his understanding and infuse new thoughts and ideas.

Summary
The first two lessons (Module-1) provide an introduction to the course. Lesson-1 explains the meaning of the individual terms: industry, automation and control and their interrelations. It also explains the necessity and importance of automation in modern day industries. Various types of processes and different classes of automation are also discussed in this lesson. Lesson-2 discusses the architecture of an industrial automation process. The functionalities of various levels of automation have been explained in this lesson. Basic sensors, actuators and the primary control loops form level zero and level one of automation, while the higher levels are supervisory and managerial level control. Constant communication and steady flow of information among the various levels is essential for efficient running of the automation process. These issues have been elaborated in Lesson-2. Lessons 3 to 10 discuss the basics of sensing schemes used in industry. Accurate measurement of different process variables is very important for the successful operation of an automated process. These variables are to be measured and according to their deviations from the set points, the input variables are manipulated. These variables are mainly, temperature, pressure, flow, level, speed, displacement, pH etc. Different sensing schemes used for measuring these process parameters have been elaborated in these lessons. A measuring system is characterized by not Version 2 EE IIT, Kharagpur 4

only the transducer used but also the measuring circuit used for signal conditioning and processing. The signal received from the transducer an electrical signal of very small strength. This signal has to be amplified, filtered and processed before it is transmitted for further control and data storage. The presence of noise in the measuring signal is inevitable in an industrial environment, but it should be tackled properly so that its effect is minimized. The accuracy and resolution of a measuring system are also limited by the noise level present. All these aspects have been addressed in module –2 (Lessons3-10) of this course. Lesson 3 discusses about the measurement system specifications. The different terms used for specifying a sensing system have been introduced. The differences between static and dynamic characteristics have also been elaborated. Lesson 4 discusses about different techniques used for temperature measurement and the compensation schemes used. The lesson concentrates on three major types of sensors: RTD, Thermocouple and Thermistor. The other methods of sensing (eg. radiation pyrometric technique) could not be discussed in this lesson. Lesson 5 discusses different sensing schemes for force and pressure measurement, while methods of displacement and speed measurements have been discussed in Lesson 6. Displacement and speed /velocity sensing is very useful in positioning of tools and other objects. The measuring scheme can be intrusive (where the sensor is directly connected to the moving object; or it may be nonintrusive, where the measurement has to be carried out without any direct contact between the sensor and the moving object. Optical and electromagnetic methods of measurement find wide use under this situation. The principles of operation of such sensors have been discussed in Lesson-6. Another important area of industrial measurement is flow measurement of flow. In fact most of the controls in process industries are achieved by maneuvering a flow control valve to control the flow rate of a stream. In this way temperature, pressure or level of a tank is controlled. But the fluid may be liquid or gaseous, the measurement requirement may be volumetric or mass flow rate, or we may need to measure the total amount of flow. All these aspects have been discussed in Lesson-7 with particular emphasis towards measurement of liquid flow rate through obstruction type meters. Similarly, the principles of measurement of pH, humidity and levelanother three important parameters used in process industries have been elaborated in Lesson-8. Lesson-9 discusses about the signal conditioning circuits. Whatever the physical parameters measured, it is convenient to convert the signal into form of a small voltage or variation of resistance, inductance or capacitance. Suitable circuits are needed to convert the signals to electrical voltage up to a desired level. Unbalanced d.c. and a.c. bridges are used for resistive, inductive and capacitive sensors. Sensitivity and linearity of the bridge are two major issues for designing the bridges. These aspects, along with different amplifier circuits have been elaborated in this lesson. Lesson-10 deals with different types of errors in a measurement system those are encountered commonly. The methods for estimating the error of a system made of several blocks in cascade have also been discussed. Another important issue discussed in this lesson is the calibration technique. Every measurement system has to be calibrated in regular frequency against some standard measuring instruments at different points of operation. Different methods of calibration and their adjustment techniques have been elaborated. Module-3 of this course comprising of seven lessons (Lesson-11 to 17) discusses on process control. Before controlling a plant or process, we must know about the process and its dynamics. Typical features of an industrial process have been discussed in Lesson-11. Presence of time Version 2 EE IIT, Kharagpur 5

delay and disturbance in a process unless properly understood would cause deterioration of the performance of closed loop system. The basic reasons behind the presence of these two parameters have been elaborated with simple examples. Examples of SISO (single input single output) and MIMO (multiple input multiple output) systems have also been provided. Another important aspect discussed in this lesson is the linearisation technique. Common processes are nonlinear in nature. But most of the elegant mathematical tools for design and analysis of control systems assume linear system behaviour. In order to achieve a compromise, nonlinear systems are often linearised over an operating point and detailed design and performance studies are carried out. The linearisation technique has also been elaborated in Lesson-11. The next three lessons (Lesson12-14) are devoted to PID controllers. PID controllers are the most popular among all industrial controllers. The effects of the individual P, I and D elements on the controlled output have been elaborated in Lesson-12. A general guideline has also been provided for selection of P, PI, PD or PID controllers. But the major problem is how to select the proportional gain, integral time and derivative time of a PID controller. They are very much process dependent and improper selection may often lead to instability and deterioration of performance. The tuning is mainly carried out after performing some experiments with the process and the controller. Various tuning rules have been discussed in Lesson-13. There are also several issues (such as smooth transition from manual to auto mode and integration windup) those need to be properly addressed before putting the PID controller in action. These issues along with different schemes for implementing PID control actions have been discussed in Lesson-14. Though PID controllers constitute of the main building block of a controller, there are several control actions those are needed for particular types of systems in order to improve the performance. These schemes, such as, Feedforward Control, Cascade Control, Ratio, Predictive control, Split range and Override control have been discussed in Lessons-15-17. Module-4 (Lessons 18-22) discusses about Programmable Logic Controllers. The control scheme used here is sequential in nature, where one operation follows another. Unlike continuous control schemes discussed in Module-3, these control strategies are open loop, logical in nature and the actuations are on-off type. Programmable Logic Controllers (PLCs) are used to generate the preprogrammed sequence of operations. Lesson-18 provides an introduction to sequence control, giving an example of an industrial use of sequential control. It also provides a lucid background on evolution of PLCs. Typical architecture of a PLC system is also discussed in Lesson-18. This is followed by description of software environment and programming of PLCs in Lesson-19. Given a desired sequence of operation, how to generate a Relay Ladder Logic (RLL) diagram has been explained. Typical switches, timers and counters used for this purpose have also been discussed. Lesson-20 discusses an approach of formal modeling and structured RLL programming. It describes a systematic method for designing the sequential control in a framework of finite state machines and the methodology for developing sequence control programs. Lesson-21 discusses about Sequential Function Chart- a standard method for generating sequential control for complex industrial systems. This is followed by Lesson-22, where the hardware of a PLC has been discussed. Modern manufacturing and machining process often demands precision positioning of tools. Computer Numerically Controlled (CNC) machines are used for providing translational and rotational motions of the tools in a preprogrammed fashion. Module-5 (Lesson-23 and 24) gives an introduction of operation of CNC machines and their control and drive schemes. Version 2 EE IIT, Kharagpur 6

Module-6 deals with different types of actuators used for generating final control actions. Hydraulic and Pneumatic Controllers, Servovalves and actuator systems have been discussed in Lessons 26-30. Lesson-25 describes different types of flow control valves, their construction and characteristics. Electrical machine drives constitute a major part of the actuation system used for automation of modern industries. Different types of electrical drives have been discussed in Module-7. Lesson 31 describes methods of flow control using industrial fans and pumps. It also shows how considerable energy can be saved using variable speed drives. The construction and principle of operation of step motors have been discussed in Lesson 32. Lessons 33-35 discuss adjustable speed drives using d.c. motors, BLDC drives and induction motor drives. Module-8 is devoted towards discussions on industrial embedded and communication systems. A Real Time Embedded System (RTES) is a computational subsystem within an overall system built to discharge a specific industrial task in real time. The typical characteristics of an RTES and its architecture have been elaborated in Lesson 36. Lesson 37 describes typical real time operating systems for such embedded systems. Lesson 38 deals with the communication systems for networking the field devices via Fieldbus. It discusses the protocol, architecture and synchronization issues of a Fieldbus system. Finally in Lesson-39 of the concluding module (Module-9) discusses hierarchy in a production control and management system. Different levels of automation, responsibility of each level and communication among each levels have been discussed. The functionality of a distributed control system has been elaborated. Lastly, the typical example of a large rolling mill automation system has been taken and the functions of lower levels of automation have been explained. Discussions on Process Control in such a course on Industrial Automation and Control must strike a balance between theory and practice. Keeping this in view, the exposition in the area of Control, which is very rich in theory, has focused towards practical features of popular control schemes, such PID controllers, feedforward and cascade control. Note that the latter two are extensions over the standard unity feedback control configuration that an undergraduate student is exposed to, in her first course on control. However, one should mention here that some of the more advanced control strategies, such as Model Predictive Control, have not been included. The interested reader can locate (without a great deal of effort, using the internet) sufficient references and material on these topics. Among some of the technology areas significant to industrial automation that have been left out or treated only briefly, are: • • • • • Process Monitoring, Fault Detection and Diagnosis Scheduling and Performance Analysis Sensor Fusion, Signal Estimation and Virtual Sensing Industrial Robotics and Material Handling Systems Manufacturing Quality Assessment and Control Version 2 EE IIT, Kharagpur 7

• •

Industrial Communication and Data Management and Computing Systems Embedded Systems.

Concluding Remarks
The authors have tried to uncover the different aspects of the subject “Industrial Automation and Control”. However catering the need of such an interdisciplinary subject is not easy. The pace and depth of presentation was also not even. But the basic motivation of the course was to give the reader an exposure to the actual working of an industrial automation system. It is intended that if the reader is acquainted with the workings of an industry or if he visits and works in a industry he would be able to appreciate the function of the various building blocks of an automation system and develop a comprehensive knowledge about its complexities and challenges. Lastly feedback on the content of the course and suggestions on further improvement of the contents would be highly appreciated by the authors. The readers may send their comments by email to the following addresses: 1. smukh@ee.iitkgp.ernet.in 2. ssen@ee.iitkgp.ernet.in.

Version 2 EE IIT, Kharagpur 8