Transducers :

UNIT 3 TRANSDUCERS : APPLICATIONS AND Applications and

Selections
SELECTIONS
Structure
3.1 Introduction
Objectives

3.2 Application of Proximity Switch

3.3 Application of Photoelectric Sensor
3.4 Sensor Array
3.5 Wrist Sensors
3.6 Compliance Sensing
3.7 Range Sensing
3.8 Guidelines for Selection
3.9 Summary
3.10 Key Words
3.11 Answers to SAQs

3.1 INTRODUCTION

The performance of the mechanical system can be greatly enhanced by providing suitable
sensors. In Unit 1, you were introduced to different type of sensors. In Unit 2, you studied
different characteristics of the sensors. In the present unit, we are going to study how
these sensors can be employed in mechanical systems. With the advent of mechatronics,
there has been a virtual explosion of sensor design, ranging from tactile (or touch) sensors
to proximity sensors, and from sensing of sound to vision. Entire conferences and books
are devoted to the subject. Clearly, only a selection can be presented here.
Objectives
After reading this unit, you should be able to
• know about the application of different sensors, and
• select a suitable type of sensor when you know the application.

3.2 APPLICATION OF PROXIMITY SWITCH

Proximity switch is a non-contact sensor. Due to this it lasts longer as it is not subjected to
mechanical forces. It is advantageous to use this sensor in hazardous applications where
remote sensing is desirable. However the area between object and the sensing head of
the proximity sensor should be neat and clean. The inductive type of switch is used for
metallic objects and the capacitive proximity switch is used for non-metallic objects as
well.
Figure 3.1(a) shows a capacitive proximity sensor in counting cans in food industry. In
Figure 3.1(b) a proximity switch is detecting the presence or absence of a metallic cap on
passing jars of a food product. This is a good application for an inductive proximity switch
since a jar without a cap will not be sensed, as it would be by a capacitive proximity
switch. The function of the capacitive proximity switch is to sense the presence of a jar.
Without this sensor, the controlle r would not know when a jar without a cap has passed.
Figure 3.1(c) shows an application of proximity switch in petroleum industry. When the

37
Mechatronics valve is open the proximity switch A is activated and when the valve is closed the
proximity switch B is activated.

Figure 3.1 : Applications of the Proximity Switch (a) Capacitive Proximity Switch Counting Canes,
(b) Inductive Proximity Switch Detecting Presence of Metal Cap,
(c) Application in Petroleum Industry to know the Valve Position

Figure 3.2 shows another interesting application of both inductive and capacitive proximity
switches. In certain industry two types of part, one made of metal and another made of
plastic are manufactured. The machines that produce the parts dump them on a main
conveyor. When the parts move on the conveyor, they first face an inductive proximity
sensor. If the part is made of metal, it is detected by this switch and is diverted to the
diverter lane A by some actuator. If the part is made of plastic, it moves ahead and is
detected by a capacitive proximity switch. It is then diverted to lane B.
Ultrasonic transreceivers can also be used for proximity measurement. The time delay
between the transmitted and the received pulse is a measure of the distance of the object
from the sensor.

38
 Transducers :
Applications and
Figure 3.2 : Application of Proximity Switches to Sort Out Metallic and Non -metallic Objects Selections

3.3 APPLICATION OF PHOTOELECTRIC SENSOR

Photoelectric sensors are designed for use in four basic sensing modes : opposed,
retroreflective, diffuse, and convergent. The appropriate sensing mode depends on the
application.
Figure 3.3(a) shows the opposed sensing mode, sometimes referred as the transmitted
beam mode. In this mode, the emitter and receiver are positioned opposite to each other.
When an object passes through the beam, interrupting it, the base of the phototransistor
receiver is de -energised. In the opposed sensing mode the object must be large enough to
cover the entire beam width. Otherwise, some light will always be incident on the receiver
and the object will not be detected. For that reason, manufacturers of photoelectric
sensors provide a control on emitter to allow a user to aperture the light down to a fairly
narrow beam. Although the opposed sensing mode is considered very reliable, it requires
the expense of purchasing separately housed emitter/receiver pairs with additional wiring
considerations. This is avoided in the retroreflective mode.
Figure 3.3(b) illustrates the retroreflective sensing mode. In this mode the emitter and
receiver are housed in the same package. The light beam is established by bouncing it off
a special target, which returns it to the receiver. Again, the principle is that an interruption
of the light beam indicates the presence of an object.
Although the retroreflective sensors are more economical, there are some additional
considerations. Objects that have a highly reflective surface may return the light beam
and pass undetected. Retroreflective sensors provide much less light intensity and their
effectiveness can deteriorate when used in an environment where dirt collects on the
reflector. On the other hand, when the target is large and non-reflective and the
environment is relatively dirt free, it offers a good solution.
Figure 3.3(c) illustrates a diffused sensing mode. This mode has the emitter and receiver
housed in the same package. In diffused sensing, light is emitted at some wide angle and
there is no reflective target. Instead, the return light comes from the object itself, which
must have a reflective surface. The receiver will pick up some portion of the returned
light, which biases the phototransistor base. The logic is opposite to the opposed and
retroreflective sensing modes; the presence of an object makes the beam rather than
breaks it.

39
Mechatronics Figure 3.3 : Various Sensing Modes of a Photoelectric Sensor Along a Conveyor (a) Opposed
Sensing Mode, (b) Retro-reflective Sensing Mode, (c) Diffused Sensing Mode, (d) Convergent
Sensing Mode

Convergent beam sensing is illustrated in Figure 3.3(d). In this mode, the emitter and
receiver are housed in the same package. The emitter light is focussed at a specific point,
which limits the sensing depth of field. An object passing through that range will reflect
light back to the receiver, which is tuned to a specific light requirement based on the
reflectivity of the object. The importance of the convergent sensing mode is that a user
can focus detection on a very specific area. For example, if it is important to check an
assembly moving along a conveyor for the presence of a part in a specific place, the
convergent sensing mode is a good choice.
Photoelectric sensors have the capability of being used over a fairly long distance. For
example, distances over 35 meters are possible in opposed sensing mode.

3.4 SENSOR ARRAY

Another configuration for using photoelectric sensors is the sensor array, shown in Figure
3.4. The sensor array provides several binary inputs, which are scanned by the controller
to determine which are on and which are off. As obje cts traverse the field of view, the
pattern of obstructed light can be used to identify the size or orientation of an object. A
typical use is to sort objects by size on a conveyor.

Figure 3.4 : An Application of Sensor Arrays

SAQ 1
(a) What is the main advantage of a capacitive proximity switch over the
inductive proximity switch?
(b) Describe four different sensing modes for photoelectric sensors.

3.5 WRIST SENSORS

Wrist sensors are used in robot wrist. The purpose of a force sensing wrist is to provide
information about the three components of force and three moments being applied at the
40 end of the arm. Based on sensory information and calculations carried out by the robot
controller to resolve the forces and moments into the six components, the robot controller Transducers :
Applications and
can obtain the exact amount of forces and moments being applied at the wrist, which can Selections
be used for a number of applications.
In most of the wrist sensors, a collection of beams is oriented in such a way that their
bending within a rigid frame may be used to deduce force and moment information.
Forces and moments by their very nature cannot be measured directly. One way of
measuring is through comparison, by using an apparatus to match an unknown force, such
as the weighing of an object on a scale. Another method is the deduction of force as a
result of the deformation of some transducer that is in contact with the force-producing
element. The standard transducer is a strain gauge.
One particular type of wrist sensor is the Maltese cross, used in medieval times as an
emblem by the knights of Malta, consisting of four arrowheads pointing toward a common
center. The concept of a Maltese cross was first used in a sensor developed by V.
Scheinman. Scheinman’s company, Vicarm, Inc., which later acquired by Unimation,
produced a refined version for Jet Propulsion Laboratories. The term Maltese cross is an
excellent mnemonic device and has helped in publicising the type of sensor shown in
Figure 3.5.

Figure 3.5 : The Maltese Cross : (a) General Configuration Machined from a Single Piece of
Aluminium, (b) Raw Data and Force and Moment Resultants

The sensor shown in Figure 3.8 is machined from a single piece of aluminium.
Measurements are obtained from pairs of strain gauges mounted on the four sides of each
of the four spokes (slender beams). The spokes are attached to the relatively flexible rim
on one side and to the square hub at the center, approximating a simple support and a rigid
support respectively.
Each pair of strain gauges – eight pairs in all – is connected to a potentiometric circuit like
the one shown in Figure 3.5. Insofar as the support conditions given previously are
satisfied, we may assume that the deformations are uncoupled and the output voltages v i
of the potentiometric circuits are proportional to the corresponding forces wi at the end of
the spokes, as shown in the figure. These raw data are related to the orthogonal force and
moment component via the matrix equation
F = Rw . . . (3.1)
where
F=(Fx Fy Fz Mx M y Mz)T and w = ( w1 w2 . . . w8)T

41
Mechatronics and where 6 × 8 matrix R = [rij] is called the resolved force matrix. When the readings
are uncoupled, as assumed previously, the matrix is sparse, since we have
Fx = r13 w3 + r17 w7 . . . (3.2)

for example, with no other contributions to Fx.

3.6 COMPLIANCE SENSING

Rigid end effectors are not preferred for certain applications. For example, when a
flexible hand is required to successfully ins ert a pin into a hole, the robot finger may
experience a small lateral error in some direction. Further, it is possible that small
orientation errors ∆θx and ∆θy may exist about the x and y axes. Figure 3.6 illustrates
orientation and lateral errors.

Figure 3.6 : (a) Orientation and (b) Lateral Errors

The pin insertion problem is considered in detail in the following. If the hole is provided
with a countersink or alternatively if the end of the pin is bevelled at its end, the pin can be
partially inserted into a hole, but cannot be easily forced in due to the lateral and
orientation errors (∆x, ∆y, ∆θx, ∆θy). Large forces measured by the force transducers
indicate that the job got stuck and further help is necessary. A flexible robot hand is
provided with displacement transducers along the x and y directions at two places
separated vertically by a distance S (Figure 3.7). The two sets of transducers read zero
when the hand is free. However, when the job is stuck awkwardly, the readings (x1, y1, x2,
y2) would help the robot determine the lateral offsets ∆x, ∆y and the angular tilts, ∆θx, ∆θy.
From Figure 3.7, it is clear that :
∆x = x2 + ( x1 − x2 ) L / S ...
(3.3)
and

42
 ∆y = y2 + ( y1 − y2 ) L / S ... Transducers :
Applications and
(3.4) Selections

The orientation errors about the x and y axes are given by :

y1 − y 2
sin ∆ θ x ≈ ∆ θ x = ...
S
(3.5)
and
x1 − x 2
sin ∆ θ ≈ ∆ θ y = ...
S
(3.6)

Thus,

∆x  L/ S 1−L/ S 0 0  x1 

∆y  
−  x2 
  = 0 0 L/ S 1 L / S

∆θx   0 0 1/ S −1/ S  y1  ...
     
∆θ
 y 1/ S −1/ S 0 0  y2
(3.7)
The robot control software takes the information [∆x, ∆y, ∆θx, ∆θy] and moves the six
joint motors such that the lateral and orientation errors are made zero.

43
Mechatronics

Figure 3.7 : Measurement of Orientation and Lateral Errors

3.7 RANGE SENSING

Light beams are sometimes used for locating objects for robots. A triangulation procedure
is frequently employed. A narrow beam of light is made to sweep uniformly the scene
under observation. This is done by rotating the source S about the z-axis at a uniform rate.
A detector is mounted rigidly on a sliding base as shown in Figure 3.8. The source
detector pair can be slided along the x-direction as well as the z-direction in a parallel
manner. The receiver detects the light scattered by the object. However, it detects only
the light collimated in the y-direction. It is clear that only when y = B tanθ, the response at
the detector is a maximum. Since θ (t), the tilt of the source is known at any given time
and B is fixed, y can be easily determined. The contour of the object can be obtained by
sliding the source detector pair in the x direction and obtainingθ, and hence y, for a fixed z.
Then z is changed by a small step and the contour is determined again. The envelope of
all these contours gives an idea regarding the shape of the surface of the object.

Figure 3.8 : Range Measurement

For determining the range and orientation, a structured lighting procedure is sometimes
employed. A flat sheet of light or a series of flat sheets of light give rise to structured
illumination. A single flat sheet of light can be used for illuminating an object and
estimating the range by using a TV camera (Figure 3.9). The TV camera plate is
discretized to when m× m pixels, m is ordinarily equal to 256.

44
 Transducers :
Applications and
Selections

Figure 3.9 : Illumination Using a Light Sheet

The calibration is done in the following manner. A vertical object of height Dc is placed on
the table and illuminated by a light sheet such that D c, top of the object, is seen at the
centre of the camera plate (Figure 3.10). Let d be equal to half the width of the camera
plate which corresponds to m/2 pixels. Let the top of another object of height Do throw its
image at the extreme edge of the camera plate. Then the following relations hold
good :

αc = tan−1 [Dc / B]
αo = tan−1 [Do / B] ...
φ = αc − αo
(3.8)
Now an object of unknown height Dk is placed before the camera whose top is imaged at
a distance of dk from the edge of the camera plate; i.e. the Kth pixel from the edge. It is
assumed that the foot of all the three objects are placed on the same point in the x − y
plane at a distance of B from the centre of the lens.
It is clear that d is proportional to m/2 and dk is proportional to K.
dk /d = K / (m/2) = 2K/m

or, dk = 2Kd/m
. . . (3.9)
clearly tan θk =(d − dk) / λ, where λ is the focal length of the camera.

Now θko = αc − αk

where αk = tan−1 (Dk / B) ...

(3.10)
From Eqs. (3.9) and (3.10), it is clear that
(d − dk) [d − (2Kd/m)]
tan θk = =
λ λ
( md − 2 Kd )
= . . . (3.11)
λm

In Eq. (3.11) d, m, λ are known constants and K is determined by the computer by actual
counting. So, from the measured K, θk can be computed using Eq. (3.11). On the object
side,
Dk = B tan αk
= B tan (α c − θ k ) ...
(3.12)
Hence the height of the object (= D k) can be easily computed using Eq. (3.12).

45
Mechatronics

Figure 3.10 : Calibration Scheme Using Height Measurement

A simple laser range meter is discussed in the following part of the section. This type of
meter finds application in mobile robots. The arrangement of a laser range meter is shown
in Figure 3.11. A continuous laser beam is passed through a semi-silvered mirror which
gets partially reflected to a detector D1 . The part of the beam passing through the glass
plate intercepts an object and the reflected beam from the object on meeting the mirror
gets deflected to another detector D2.
The path difference between the two beams reaching D1 and D2 is clearly 2D. The phase
shift between the transduced signals at D 1 and D2 is a measure of the distance D. We can
show that
(θλ / 360) 0 = 2D .
. . (3.13)
where λ is the wavelength of the laser and θ is the phase shift in degrees.

Figure 3.11 : A Laser Range Meter

Even if the phase shift θ could be measured accurately, the operating range of the system
would be very small because it would be of order of the wavelength of the laser
46
(Figure 3.12). To expand the range, the scheme is slightly modified. Instead of a plain Transducers :
Applications and
laser beam, a modulated laser beam is used (Figure 3.13). If the frequency of the Selections
modulating wave is 10 MHz,the wavelength of the modulating signal, computed at the
velocity of light (= 3×108 m/ s) is obviously,

3×108
λαM = = 30m .
10×106
. . (3.14)
This range is adequate for robotic applications. The measurement procedure would be to
modulate the detected signals at D1 & D2 and compare the envelopes of the transmitted
and the received signals. If the phase shift between them is found to be θM degrees, then
θM λM
= 2D .
360
. . (3.15)

Figure 3.12 : Unmodu lated Transmitted and Received Rays

Figure 3.13 : Modulated Transmitted and Received Rays

SAQ 2
(a) Describe the working of a robot wrist sensor.
(b) Write a short note on the methods for range sensing.

47
Mechatronics
3.8 GUIDELINES FOR SELECTION

Sensors can be categorised according to their characteristics, the physical parameter to

be sensed, and their applications. As discussed previously, the sensors could be of the
contact type like switches, probes, feelers, pressure pad etc. The non-contact type of
sensors operate at a distance and can be based on x-rays, magnetic fields, light, infra-rays
and so on.
A sensor is normally evaluated by measuring its sensitivity, linearity, range, response time,
repeatability and resolution. Five important characteristics of sensing device which should
be looked for are :
Range
This refers to the minimum and maximum change in input signal to which the
sensor can respond. The sensor should possess a wide operating range.
Response
The sensor should be capable of responding to changes in the sensed variable in
minimum time. Ideally, the response should be instantaneous.
Accuracy
The accuracy of the measurement should be as high as possible. The output of the
sensing device should be properly reflect the input quantity being measured or
sensed.
Sensitivity
It refers to the change in the output exhibited by the sensor for a unit change in
input. The sensitivity should be as high as possible.
Linearity
The sensory device should exhibit the same sensitivity over its entire operating
range.

Other vital physical characteristics based on the type of output are:

(a) electrical (voltage, current or frequency), and
(b) mechanical (pressure, level or intensity of illumination).
A sensor must be light in weight, compact and reliable. It should be possible to interface a
sensor with a microcomputer. Also important are the physical size, cost and ease of
operation. The following are the other considerations.
(i) The device should not disturb or have any effect upon the quantity it senses
or measures.
(ii) The device should be suitable for the environment in which it is to be
employed.
(iii) Ideally, the device should include isolation from receiving excess signals or
electrical noise that could give rise to the possibility of mis-operation or
damage of the sensor, circuit or system.
Sometimes no single sensor is able to perform the task of providing all needed
information. In that case a number of sensors are integrated in a manner in which the
individual data from each of the sensors (force, vibration, temperature, dimensions etc.)
are combined to provide a higher level of information. This is called sensor fusion. It has
48 been suggested that a common example of sensor fusion is someone drinking from a cup
of hot tea or coffee. Although we take such an everyday event for granted, it can readily Transducers :
Applications and
be seen that this process involves data input from the person’s eyes, hands and lips and Selections
tongue. There is a real time monitoring, through various senses, of relative movement and
positions, as well as temperature. Thus, if the fluid is too hot, the hand movement of the
cup toward the lip is controlled accordingly. The sensor fusion technology has been
applied to robot movement control and missile flight tracking, primarily because these
activities involve movements that mimic human behaviour.
SAQ 3
What are the major guidelines for the selection of a sensor?

3.9 SUMMARY

Sensors are considered to be the eyes of a mechatronic system. The different type of
sensors are available in the market. A particular sensor may be utilised for different
applications. A sensor must be selected on the basis of physical quantity to be sensed and
the type of response desired. The manufacturer’s catalogue must be referred before
selecting a particular type of sensor. The durability of the sensor, reliability, compactness
and the cost are other deciding factors for selecting a sensor.

3.10 KEY WORDS

Compliant End Effectors : These effectors(grippers) can sense the force

applied to the object being handled, using
piezo-electric devices, strain gauges, magnetic
induction, ultrasonics, and optical systems of fibre
optics and light emitting diodes.
Robot : An industrial robot is an automatic, servo-control,
freely programmable, multipurpose manipulator,
with several axes, for the handling of work pieces,
tools or special devices. Variably programmed
operations make it possible the execution of
multiplicity of tasks.
Robot Wrist : The robot wrist is the joint having three to five
degrees of freedom to orient the end-effector
(gripper) properly.
Tactile Sensing : It is the continuous sensing of variable contact
force, commonly by an array of sensors.

3.11 ANSWERS TO SAQs

SAQ 1
One of the shortcomings of the inductive proximity switch is that it can only sense
metal objects. The capacitive proximity switch, on the other hand, can sense non-
metallic objects as well.

49
Mechatronics Check your answers of all other SAQs with respective preceding text of each
SAQ.

50
 Transducers :
FURTHER READINGS Applications and
Selections

Patranabis D., (1997), Sensors and Transducers, Wheeler Publishing, Allahabad.

Boucher Thomas O., (1996), Computer Automation in Manufacturing : An Introduction,
Chapman & Hall, Madras.
Stadler W., (1995), Analytical Robotics and Mechatronics, McGraw-Hill inc., New Delhi.
Deb S. R., (1994), Robotics Technology and Flexible Automation, Tata McGraw-Hill
Publishing Company Limited, New Delhi.
Rangan C. S., Sharma G. R. and Mani V. S. V., (1997), Instrumentation Devices and
Systems, Tata McGraw Hill Publishing Company Limited.
Taylor H. R., (1997), Data Acquisition for Sensor Systems, Chapman & Hall.
Catalogues of various manufacturers.

51