Unit 5

Closed Loop Controllers

Reference
1. C.D.Jonhson, “Process Control And Instrumentation”, Chapter 9 and 10.
2. Terry Bartelt, “ Process Control Systems and Instrumentation”, Chapter 3.
 What Is Control System?
 A Control System is an arrangement of physical components
connected/related in such a manner as to command, direct or regulate itself
or another system.

 A Control System consists of subsystems and processes (or plants)

assembled to control the outputs of a process.
 Classification Of Control System
A. Depending on Objective (What to control)
1) If the aim is to maintain a physical variable at some fixed value when there are
disturbances, this is called as regulator. It is also alternatively known as Process
Control [Continuous Process Control & Batch Process Control]
Example: Temperature Control, speed-control system on the ac generators of
power utility companies.

2) The second class is the servomechanism. This is a control system in which a

physical variable is required to follow (track) some desired time function.
Example: an automatic aircraft landing system, or a robot arm designed to follow
a required path in space.

B. Depending on Configuration
1) Open loop system 2) Closed loop system [Feedback & Feed forward Control]

C. Depending on Signal
1) Analog Control System 2) Digital Control System
 Open-Loop Control System

 Process control operations are performed automatically by either open-

loop or closed-loop systems.

 Processes controlled only by set-point commands without feedback are

open-loop.

 Open-loop systems are used in applications where simple processes are

performed.

 Open-loop systems are relatively inexpensive.

Set Point Controller Process

Block Diagram of open loop control system

 Closed-Loop Control System

 Closed-loop control systems are more effective than open-loop systems.

 With the addition of a feedback loop they become self-regulating.

 Components of a closed-loop system include:

I. The primary element – sensor

II. The controlled variable

III. The measured variable

IV. The control signal

V. The final correcting element

 Block Diagram Of Closed loop Control System

SET ERROR ERROR CONTROL ACTUATOR & FINAL

POINT DETECTOR SIGNAL CONTROLLER SIGNAL CONTROL ELEMENT

ACTION
FEEDBACK

CONTROLLED
SENSOR(FEEDBACK PROCESS
ELEMENT) VARIABLE
 Process
 Complex assembly of phenomenon that refers to some manufacturing
sequence. It utilizes the resources to produce certain product.

 Many variables may be involved in such a process, some of which have to be

controlled.

 Classification of processes.
Process

Based on variables Based on operation

to be controlled
Batch(sequential) Continuous
Single variable Multi variable process process
process process
 Process Behavior
 The objective of process control is to cause a controlled variable to remain at a
constant value at or near some desired set-point.

 The controlled variable changes because of:

1. A disturbance appears
2. Load demands varies or
3. Set points are adjusted.

 Several process variables are controlled at once in a typical production

machine.

 Usually, only one individual feedback loop is required to control each variable.

 Single-variable control loops consist of the following elements:

o Measuring device
o Transducer/transmitter
o Controller
o Final Control Element
 Process Behavior Example

 Flow through the pipe

is the process.

 Fluid flow rate is the

controlled variable.

 Valve position is
the set point.

 Demand for the fluid

downstream is the load.

 Variance in upstream
pressure is the disturbance.
 Sensor (Feedback Element)

 It is the eye of the system.

 Produces output that represents the status of the controlled variable.

 The output of the sensor is called as feedback.

 Examples of sensors used in process control are—

 Thermal sensors like RTD, Thermistor, Thermocouple etc.

 Level sensors like Ultrasonic, Float, Radiation sensor etc.

 Pressure sensors like Diaphragm, Bourdon tube, Bellows etc.

 Flow sensors like Ventury meter, magnetic flow meter etc.

 Optical sensors based on LED, LASER and Photodiode, Phototransistor.

 Feedback Elements(Sensors)

Strain gauge
Type of Signals
Pressure Piezo-electric
Capacitance Transmitters
Pneumatic
Bourdon Tube
3-15 PSI
Mechanical Floats
Pressure Transmitter Electrical
Level Guided Wave
Weight (load cell)
Level Transmitter Current
Ultrasonic
4 – 20 mA
pH Static Pressure
Differential Pressure 0 – 20 mA
Cell 10 – 50 mA
Humidity
Head meters Flow Transmitter Voltage
(orifice, venturi) 0–5V
Flow Coriolis Mass, Temperature
1–5V
velocity, Transmitter
Density 0 – 10 V

Speed Thermocouples Digital

RTDs / Thermistors ON/OFF
Temperature Filled Systems Field Bus
Bi-metallic ModBus
ProfiBus
 Error Detector

 Compares set point(reference signal) with the feedback signal.

 Produces error signal for the controller.

 Error = Set point – feedback signal.

 Examples of devices used as error detector are—

- OP-AMP based differential amplifier for analog signals.

-Comparison soft wares for digital signals.

 Controller
 The controller is the brain of the system.

 Receives error signal and develops output that causes the controlled variable to become
equal to the set point value.

 Examples of controller PLC, Microprocessor, OP-AMP based controller.

 Different modes of controller are—

Controller modes

Discontinuous mode Continuous mode

● Two position controller ● Proportional controller
● Multiposition controller ● Proportional Integral controller
● Floating controller ● Proportional Derivative
controller
● Proportional Integral Derivative controller
 Selection of a Controller
 Controllers are designed to operate by using different control modes. Each of
these modes has specific characteristics to provide different types of control
actions.

 These control modes are:

1. On/Off
2. Proportional
3. Integral
4. Derivative

 The mode or combination of modes which is selected by the designer

is determined by the requirement of the process.
 On-Off Controller (2 position controller)

 Used for slow acting operations

where lag is unavoidable.

 Final correcting element is either

fully-on or fully-off.

 The primary drawback of on-off

control is the rapid switching of the final
control element.

 On-off differential or hysteresis

is programmed into the controller
to reduce cycling.

 Dead band refers to the differing levels

at which a controller switches on and off
 On-Off Controller (2 position controller)
 The analytical equation for On-Off controller is given as
P= 0 % Ep < 0
P= 100 % Ep > 0

 Applications of On-Off controller: It is best adapted to large scale systems

with relatively slow process rates. Examples of such process are
1. Room heating or air conditioning system
2. Liquid bath temperature control
3. Level control in large volume tanks.
 Continuous Control

 On/Off control is acceptable for process where the variable is set

between two limits.

 For processes where the variable needs to be kept at particular set point
level, proportional control is used.

 Proportional action can be accomplished in two ways:

– Time Proportioning Method
– Amplitude Proportional Method
 Time Proportioning Control
 Is a method whereby the output of the controller is continually switched
on and off. This method is also called as PWM(Pulse Width Modulation).
 Here the ratio of On time (Ton) to Total time (Ton+ Toff) is called as duty
cycle is varied as per the changes in the feedback signal.
 On versus off times are varied dependent upon process requirements.
 Example: Speed control of DC Motor .
 Amplitude Proportional
 Most common technique to produce a proportional signal.

 The control signal is proportional in amplitude to the error signal.

 The signal may be amplified and the amplification may be referred to as

proportional gain and proportional band.
 Proportional Control
 Smooth relationship exists between the controller output and the error.

 P= Kp Ep + Po
Where Kp- Proportional gain between error and the controller output
Po- Controller output without error

 The range of error to cover 0% to 100% controller output is called as proportional

band.
PB = 100/ Kp

 Disadvantage of proportional controller is offset or SSE or residual error.

 The offset error limits use of proportional mode to only a few cases particularly
where manual reset of the operating point is available to reset the offset.

 It is generally used in process where large load changes are unlikely or with
moderate to small process lag times.
Characteristics of Proportional Controller
Offset Error
 Integral Control
 Because of the introduction of offset in a control process, proportional
control alone is not used. It is often used in conjunction with Integral
control.

 Offset is the difference between set point and the measured value after
corrective action has taken place.
 Integral Control
The offset error of the proportional mode occurs because the controller can not
adapt to changing external conditions i.e. changing loads. In other words the zero
error output is a fixed value.

 The Integral mode eliminates this problem by allowing the controller to adapt
to changing external conditions by changing zero error output.

 Integral action is provided by summing the error over time, multiplying that
sum by a gain and adding the result to present controller output.

 Integral mode controller action the rate of output change depends on the error.


Characteristics of Integral Controller
 Derivative Mode
 For rapid load changes, the derivative
mode is typically used to prevent
oscillation in a process system.

 The derivative mode responds to the

rate of change of the error signal rather
than its amplitude.

 Derivative mode is never used by

itself, but in combination with
other modes.

 Derivative action cannot remove offset.

 The Derivative mode must be used with great care and usually with a small
gain, because a rapid change of error can cause very large, sudden changes of
controller output which can lead to instability.

 Derivative controller is not used alone because it provides no output when the
error is constant.

 It is also called as rate controller or anticipatory control as it can take an action

in advance depending upon the rate of error change.
Characteristics Of Derivative Controller
 Limitations of Derivative Control
 The derivative mode acts upon the rate of error signal change and it may
cause unnecessary upsets.

 It tends to react to sudden set point changes and will amplify noise.

 The control algorithm can be altered so that derivative acts on the

measurement and not on the error. This will reduce upsets.

 Excessive noise and step changes in the measurements can be corrected by

filtering out any change in the measurement that occurs faster than the
maximum speed of response of the process.

 DCS system provides software with adjustable filters for each variable.

 The time constant of these filters is usually adjusted from 0 to 100 seconds.

 In analog system Inverse Derivative control mode is often used.

 Inverse Derivative Control Mode
 This control action is used on fast processes. The inverse derivative mode is
opposite of the derivative mode.

 While the output of Derivative mode is directly proportional to the rate of

change in error, the output of inverse derivative mode is inversely proportional to
the rate of change in the error.

 Inverse derivative is used to reduce the gain of the controller at high

frequencies and is useful in stabilizing the control loop.

 The dynamic gain of the derivative function is selected to be 0.5 or less.

 The gain of the inverse derivative controller decreases from the proportional
gain at low frequency to the limiting value of the proportional gain divided by this
factor at high frequency.

 A proportional plus inverse derivative controller provides high gain to minimize

offset at low frequency and low gain to stabilize the loop at high frequency.
 Inverse derivative can be added to PI Controller to stabilize the loops
requiring very low proportional gains for stability.

 Inverse derivative should only be added when the loop is unstable at the
minimum gain setting of the PI Controller.

 It is available in the separate unit can be added to the loop when stability
problem occurs.

 The addition of inverse derivative when proportionally tuned has little effect
on the natural frequency of the loop.
 Proportional Integral (PI) Control
 Combines proportional and integral mode together and eliminates the offset
inherent in proportional controller.

 However makes the action sluggish and increases the response time.

 Another disadvantage of this system is that during start up of the batch

process the integral action causes a considerable overshoot of the error and the
output before settling to the operation point.
Characteristics of PI Controller
 Proportional Derivative (PD) Control
 It involves the serial (cascaded) use of the proportional and derivative modes.

 The analytical expression for this mode is given as

 This controller cannot eliminate the offset of proportional controllers , however it

can handle fast process load changes.
Characteristics of PD Control
 Proportional Integral and Derivative (PID) control
 One of the most powerful but complex controller mode operations combines
the proportional, integral and derivative modes.

 The system can be used for virtually any process condition.

 The analytical expression is

 This mode eliminates the offset of the proportional mode and still provides
fast response.
PID controller characteristics
Control Mode Summary
 Electronic Controllers
1. On Off Controller with dead band

 Here VH = Vsp and VL = Vsp- (R1/R2) Vo

 Proportional Controller
 We know that for proportional mode
p = Kp Ep + Po

 For implementation of electronic controller we have

Vout = Gp Ve + Vo
Where GP = R2/R1 = Gain of the controller
Application of proportional control for Robot Arm control
 Proportional Integral Controller
 We know that the control mode equation for this mode is given as

 For electronic implementation we have

 Where Proportional gain = R2/R1

Integral gain = 1/ R2.C
Proportional Integral Controller
Application of Proportional Integral Controller to Robot arm control
 Proportional Derivative Controller
 We know that for the PD Control mode the equation is given as

 For electronic implementation we have

 Where Proportional gain = R2/(R1+R3)

Derivative gain = R3.C
Proportional Derivative Controller
Application of Proportional Derivative Controller for Robot Arm Control
 Proportional Integral Derivative Controller
 We know that for PID the analytical expression is given as

 For electronic implementation we have

 Where Proportional gain= R2/R1

Integral gain = 1/RI. CI
Derivative gain = RD . CD
Proportional Integral Derivative Controller
Application of Proportional Integral Derivative Controller for Robot Arm Control
 Digital PID Controller
 
 t n 1 D 
pn  p  K c en   ek   en  en 1 
  I k 1 t    

  
 I D 

finite difference approximation

where,
t = the sampling period (the time between successive
samples of the controlled variable)
p n = controller output at the nth sampling
instant, n=1,2,…
e n = error at the nth sampling unit

velocity form -
(pn)- incremental change
 Final Control Operation

Signal Actuator
Control signal
conditioning (motor)

Final control
Process element
(valve)
 Tuning the Controller

 Fine-tuning is the process to optimize the controller operation by

adjusting the following settings:
Gain setting (proportional mode)
Reset rate (integral mode)
Rate (derivative mode)

 Three steps are taken when tuning a systems

1. Study the control loop
2. Obtain clearance for tuning procedures
3. Confirm the correction operation of the system components
 Trial-and-Error Tuning

Does not use mathematical methods, instead a chart recorder is used

and several bump tests are made in the proportional and integral modes.

Trial-and-error tuning is very time consuming and requires

considerable experience on the part of the technician or operator.

Ziegler-Nichols Tuning Methods

 Two formal procedures for tuning control loops:

1. Continuous cycling method
2. Reaction curve method
 Continuous Cycling Method
The continuous cycling method analyzes the process by forcing the
controlled variable to oscillate in even, continuous cycles.

The time duration of one cycle is called an ultimate period. The

proportional setting that causes the cycling is called the ultimate
proportional value.

These two values are then used in mathematical formulas to calculate

the controller settings.
 Ziegler-Nichols Reaction Curve Tuning Method
This method avoids the forced oscillations that are found in the continuous
cycle tuning method.

Cycling should be avoided if the process is hazardous or critical.

This method uses step changes and the rate at which the process reacts is
recorded.

The graph produces three different values used in mathematical calculations

to determine the proper controller settings.

 This method is applicable only to processes with self regulation

characteristics.

 From the response curve the following parameters are calculated

L: Lag time in minutes
Δ Cp : Controlled variable change in %
T : Process reaction time in minutes
N = Δ Cp / T = Process reaction rate in % / min.
Ziegler-Nichols Reaction Curve Tuning Method
 The calculations for Ziegler Nichols Process reaction method are

Mode Kp Ti = 1/ Ki Td = 1/ Kd

Proportional Δ P/ NL

Proportional 0.9 Δ P/ NL 3.33 L

Integral (PI)

Proportional 1.2 Δ P/ NL 2L 0.5 L

Integral
Derivative (PID)
 Frequency Response Method
 This method involves use of Bode Plot for the process and control loops.

 The method is based on the application of the Bode plot stability criterion and
the effects that the proportional gain, integral time and derivative time have on
the Bode plot.

 Bode plot stability criterion

1. If the phase is less than 140 degrees at the unity gain frequency the system is
stable. This then is 40 degrees phase margin from the limiting value of 180
degrees .
2. If the gain is 5 dB below unity when the phase lag is 180 degrees the system
is stable. This is then 5 dB gain margin.

 Tuning : The operations of tuning using frequency response method involve

adjustments of the controller parameters until the stability is proved by the
appropriate phase and gain margins.
Proportional Action : Multiplies gain curve by a constant and no effect on phase.
Integral Action : Integral gain= Ki/ω and Integral phase = - 90 degrees (lag)
Derivative Action : Derivative gain = Kd*ω and Derivative phase=90 degrees
(lead)
Bode Plot
Tuning using Bode Plot
Numerical on controller tuning:

1. In an application of the Ziegler- Nichols method, a process begins

oscillation with a 30% proportional band in an 11.5- minute period.
Find the nominal three-mode controller settings.
30% proportional band means the gain is
Pu = 100/ PB = 100/30 = 3.33
Tu = 11.5 minutes.
From the standard equations of PID tuning by Ziegler- Nichols method
Kp = 0.6*Pu= 2
TI= Tu/2 = 5.75 minutes
TD= Tu/8 = 1.44 minutes

2. In the Zeigler-Nichols method, the critical gain was found to be 4.2 and
the critical period was 2.21 minutes. Find the standard settings for
a) Proportional-mode control, b) PI control, and c) PID control.