GENERALIZED PERFORMANCE

CHARACTERISTICS OF
INSTRUMENTS
Lecture 8
Instructor : Dr Alivelu M Parimi
Example
2
IQ scores of result are normally distributed to mean of 100 and standard deviation of 13.
Find (i) Fraction of students having an IQ > 133
(ii) Fraction of students having an IQ > 90
(iii) IQ value exceeded by upper quartile (from 75% to 100%)
Solution
(i) IQ >133
z =
13
100 - 133
= 2.54, Area corresponding to z = 2.54 is 0.4945, students having IQ s 133
are 0.5 + 0.4945 =0.9945, i.e. 99.45% have IQ s 133
Fraction of students having IQ > 133 are ( 0.5 0.4945 = .0055) . i.e. 0.55% will have IQ >
133
(ii) For IQ = 90, z =
13
100 - 90
= 0.77 Area from 0 to 0.77 = 0.2799 which is same as
area for z = -0.77 to 0.
Total area 0.5 + 0.2794 = 0.7794, i.e. 77.94% will have IQ score greater than 90
(iii) Upper quartile implies area > 0.75
from Table 3.2, find z corresponding to area of 0.25 which is z
o
x x
= = 0.6745
So x x = 0.6745 x 13 = 8.76,
x = 108.76, hence upper quartile will be for students having IQ > 108.76
Gaussian distribution
3








Table 3.2: Values of Area under Gaussuain Distribution Curve
Table list areas under the
curve between zero and
various values of z
This table will have areas
from 0 to maximum 0.5
For many analytical techniques, we need to evaluate the response
of the unknown sample against a set of standards (known
quantities).
1) Determine the instrumental responses for the standards.
2) Find the response of the unknown sample.
3) Compare the response of the unknown sample to that
from the standards to determine the concentration of the
unknown.
This involves a calibration!
DYNAMIC PERFORMANCE
CHARACTERISTICS
Instruments rarely respond instantaneously to changes in the
measured variables.
Instead, they exhibit slowness or sluggishness due to factors
such as mass, thermal capacitance, fluid capacitance or
electrical capacitance.
In addition, pure delay in time is often encountered where the
instrument waits for some reaction to take place.
Such industrial instruments are nearly always used for
measuring quantities that fluctuate with time.
Therefore the dynamic and transient behavior of the
instrument is as important as the static behavior.
5
DYNAMIC PERFORMANCE
CHARACTERISTICS
The parameters characterizing the dynamic characteristics of
an instrument are:
Speed of response: It is the rapidity with which an instrument
responds to changes in the measured quantity.
Fidelity: It is the degree to which an instrument indicates, the
changes in the measured variable without dynamic
error(faithful reproduction)
Lag: It is the retardation delay in the response of an
instrument to changes in the measured variable.
Dynamic error: It is the difference between the true value of a
quantity changing with time and the value indicated by the
instrument .

6
DYNAMIC PERFORMANCE
CHARACTERISTICS
When measurement problems are concerned with rapidly
varying quantities, the dynamic relations between the input
and output are generally defined by the use of differential
equations. For most measuring instruments, generalized
relation between any particular input x and output y can be
written as:

7
a
n
n
n
dt
y d
+ a
n-1
1 - n
1 - n
dt
y d . .
+ a
1
dt
dy
+ a
0
y = b
n
n
n
dt
x d
+ b
n-1
1 - n
1 - n
dt
x d
..+ b
1
dt
dx
+ b
0
x
DYNAMIC PERFORMANCE
CHARACTERISTICS
Types of Standard Inputs
Once response of an instrument to certain standard inputs is
known, its response to more complicated inputs can be
determined using mathematical techniques.
Step input: System is subjected to an instantaneous and finite
change in measured variable.
8
DYNAMIC PERFORMANCE
CHARACTERISTICS
Linear/Ramp input: System is subjected to a measured
variable which is changing linearly with time. Like
thermometer is put in a bath whose temperature is rising
linearly.
9
DYNAMIC PERFORMANCE
CHARACTERISTICS
Pulse/Impulse input: System is subjected to step magnitude
for small time and then is brought back to its initial state.
When this small time is tending to zero, pulse becomes
impulse, which is a theoretical concept not realizable
practically as it implies in no time system is subjected to large
change. If the automobile running on road suddenly hits a pot
hole on the road, then the force experienced by shock
absorbing system may be modeled by impulse function

10
DYNAMIC PERFORMANCE
CHARACTERISTICS
Sinusoidal input: System is subjected to measured variable,
the magnitude of which changes in accordance with a
sinusoidal function of constant amplitude.
11
DYNAMIC PERFORMANCE
CHARACTERISTICS
Order of a System
The order of the system relates to the number of energy
storage elements in the measurement system.
Zero order,

First order,

Second order instrument

12
DYNAMIC PERFORMANCE
CHARACTERISTICS
Order of a System: examples
For example, if you use a mercury-in-glass thermometer to
measure temperature there is a single energy store (the heat
capacity of the mercury), thermometer acts predominantly
like a First order system.
RC and RL circuits are examples of first order system as only
electrostatic or electromagnetic energy is involved.
RLC circuit is an example of second order system as both
electromagnetic and electrostatic energies are involved.
Accelerometers and seismographs having mass-spring-damper
arrangement have two energy stores the spring (elastic
potential energy) and the mass (kinetic energy).
Thus, these devices respond like Second order systems.

13
DYNAMIC PERFORMANCE
CHARACTERISTICS
Zero Order instruments

14
a
n
n
n
dt
y d
+ a
n-1
1 - n
1 - n
dt
y d . .
+ a
1
dt
dy
+ a
0
y = b
n
n
n
dt
x d
+ b
n-1
1 - n
1 - n
dt
x d
..+ b
1
dt
dx
+ b
0
x
y =
0
0
a
b
x
Since y = Kx is algebraic, it is clear that no matter how x
might vary with time, the instrument output (readings )
follows it perfectly with no distortion or time lag of any
sort. Thus, the zero order system represents ideal or
perfect dynamic performance and is thus a standard
against which less perfect instruments may be compared.
DYNAMIC PERFORMANCE
CHARACTERISTICS
A practical instrument of a zero order type is the
displacement potentiometer shown in Figure, where a strip of
resistance material is excited with a voltage and provided with
a sliding contact. If the resistance is disturbed linearly along
length L, we may write
15
e
0
=
L
xi
E
b
= K x
i

DYNAMIC PERFORMANCE
CHARACTERISTICS
An ideal potentiometer acts like a zero order instrument.
Output reflects input faithfully without any time delay.

Ideal Operational amplifier used as an inverter is a zero order
system with a gain of (1.).

A wire strain gauge in which the change in the electrical
resistance of the wire is proportional to the strain in the wire
is also an example of zero order system.
16
DYNAMIC PERFORMANCE
CHARACTERISTICS
Step response of a zero order instrument
In a zero order instrument, y (t) = Kx (t), we have y (t) = 0 for t
< 0, and y = K for t 0. Therefore the response to the unit step
function is a step function with height K
17
Dynamic performance characteristics:
First Order instruments
18
The governing equation for the First order system is
a
1

dt
dy
+ a
0
y = b
0
x
Any instrument which follows this equation is by definition a first order instrument.
The equation can be rewritten as
0
1
a
a
dt
dy
+ y =
0
0
a
b
x
t =
0
1
a
a
, where t is the time constant.
K =
0
0
a
b
, where K is static sensitivity/gain.
Kx y
dt
dy
= + t
Dynamic performance characteristics:
First Order instruments
Example
Time constant of measurement system is determined by the
physical characteristics of the system.
A measurement system with a small time constant will
respond quicker to changes than a system with a large time
constant.
The time constant depends on the size of the energy storage
element as well as how quickly the energy can get in/out of
that storage element.
For example, mercury in glass thermometer will respond
quicker when plunged into water than it does in air (faster
heat transfer with water).
A thermometer with a large bead will respond slower than a
thermometer with a small bead

19
Dynamic performance characteristics:
First Order instruments
Step response of first order instruments
Assuming system is in equilibrium at t = 0 and then a step
input of magnitude A is applied.
20
Kx y
dt
dy
= + t
y(t) = AK (1 e

- t/t
)
Steady state value ( as t ) of y(t) = AK = y
ss
Error = e
m
= input output =
t t
= =
/ t / t
Ae ) e 1 ( A A K / ) t ( y ) t ( x
Steady state error ( as t ) = e
m,ss
= A
So
t /
,
t
ss m
m
e
e
e

=
X(s)
Y(s)
=
) s (1
K
t +

Dynamic performance characteristics:
First Order instruments
21
Nondimensional Response of First Order
System to Step Input
Non Dimensional Error Curve of First Order
System
Dynamic performance characteristics:
First Order instruments
Ramp response of first order instruments
Measuring instruments may be placed in an environment
whose temperature, pressure is increasing or decreasing in
linear (ramp) fashion.
22
X(s)
Y(s)
=
) s (1
K
t +

X(s) =
2
s
M

KM
y(t)
= -t + t + te
-t/t

Dynamic performance characteristics:
First Order instruments
23
K
) t ( y
= M ( t - t + te
-t/t
)
At larger values of t, the exponential term will tend to zero.
y = KM ( t - t ); Input = Mt, implying there is a steady state lag of t
Measurement error e
m
is given by:
e
m
= Input Output = x(t) y(t)/K = Mt M ( t - t + te
-t/t
)
e
m
= Mt - Mte
-t/t
= Mt ( 1- e
-t/t
)
Steady state error = e
m,ss
( e
m
as t ) = Mt , Transient error = Mte
-t/t

For e
m,ss
to be small, M and t should be small.
ss m,
m
e
e

= 1- e
-t/t

At steady state, measurement error is equal to the steady state error.
Dynamic performance characteristics:
First Order instruments
Impulse response of the first order system
A unit impulse function is defined as a signal which has zero
values at all times except at t = 0, where its magnitude is
infinite. In real, practical systems, it is not possible to produce
a perfect impulse to serve as input for testing. Therefore, a
brief pulse is used as an approximation of an impulse.
Provided that the pulse is short compared to the impulse
response, the result will be near enough to the true,
theoretical, impulse response
24
Dynamic performance characteristics:
First Order instruments
25
For 0 < t < T:
( D + 1) y = K x =
T
KA

Since, up until time T, this is no different from a step input of size
T
A
. For initial condition y
= 0 at t = 0, the complete solution is
y =
T
KA
( 1-e
-t/
)
at t = T, y =
T
KA
( 1-e
-T/
) (A)
For t > T, x = 0
Therefore, (D+1) y = Kx = 0
y(t) = Ce
-t/

Therefore, y(t) = Ce
-t/
(B)
At t = T, y(t) = Ce
-T/

Y(t) = Ce
-t/t
=
T
KA
t
t
/
-T/
e - 1
T
e

e
-t/

Dynamic performance characteristics:
First Order instruments
26
Dynamic performance characteristics:
First Order instruments
Response of FOS to a sinusoidal input function
The response of a system to sinusoidal input is known as the
frequency response.
If the input is sinusoidal, the steady state output is also
sinusoidal.
The output frequency equals the input frequency and the
output phase lags behind the input phase.
It is important to study the sinusoidal response as many
natural phenomena are sinusoidal in character like vibration of
guitar string, currents in oscillating circuit. In generation and
transmission of electric power and communication also
sinusoidal signals are used.
27
Dynamic performance characteristics:
First Order instruments
28
For example: if x = x
m
sin et then
y (t) =
) 1 (
) t sin( Kx
2 2
m
t e +
+ e

On comparing the input with the output, it is observed that
1. Frequency does not change in first order system.
2. There is attenuation in amplitude by a factor
) (1
K
2 2
e t +
.
3. Output lags behind the input by tan
-1
(- ).
The frequency response has two elements which can be plotted graphically
(i) The ratio of output amplitude to input amplitude plotted against frequency.
(ii) The phase lag between output and input as a function of frequency.
Dynamic performance characteristics:
First Order instruments
29
Frequency response is often specified by its
bandwidth: this is the range of frequency over
which variation in magnitude is less than 70%
(3dB).
Problem
A certain thermometer has time constant of 15 sec and an
initial temperature of 200 C. It is suddenly exposed to
temperature of 1000C. Determine the time it would take to
reach 90% of step size.
30
Solution
t = 15 sec. Step size = final value initial value = 100 20 = 80
0
C
90 % of 80
0
= 72
0
C
Therefore, total rise = 20
0
C + 72
0
C = 92
0
C
Y (t) = AK (1-e
-t/t
) + 20
(assuming K=1)
A = 80
0
C
92 = 80 (1-e
-t/t
) + 20 therefore t = 34.54 sec.
The thermometer will take 34.54 seconds to reach 90 % of the step size.
Problem
A mercury thermometer is kept in a bath of 40
0
C and bath is
heated such that its temperature increases linearly at the rate
of 5
0
C/min. Calculate the thermometer reading after 5
minutes when t is (i) 10 sec (ii) 2 min.

31
(i) t = 10 sec.
After 5 minute, actual temperature becomes 40 + 5*5 = 65
0
C. But due to error thermometer
indicates different value.
K
) t ( Y
= M (t t + te
t/t
)
t
t
=
sec 30
min 5
= 30 .
Transient term is 10 e
30
~ 0 and can be neglected

Steady state error = Mt = 5
min
0
C
* min
60
10
= 0.83
0
C
Thermometer reading = 65 0.83 = 64.17
0
C
Solution contd
32
(ii) t = 2 min,
2
5 t
=
t
= 2.5
te
t/t
= te
2.5
= 0.165
Y (t) = M (t t + t e
2..5
) + 40
Y (t) = M ( 5 2 + t e
2.5
) + 40 = 55.82
0
C
Error at t = 5 min is 65 55.82 = 9.18
0
C
Steady state error = Mt = 5x 2 = 10
0
C
Note: Smaller is the time constant lesser is the steady state error.