Measurement Errors

The measurement of any quantity plays very important role not only in science but in
branches of engineering, medicine and in almost all the human day to day activities.
<111

The technology of measurement is the base of advancement of science. The role of science
and engineering is to discover the new phenomena, new relationships, the laws of nature
and to apply these discoveries to human as well as other scientific needs. The science and
engineering is also responsible for the design of new equipments. The operation, control
and the maintenance of such equipments and the processes is also one of the important
functions of the science and engineering branches. All these activities are based on the
proper measurement and recording of physical, chemical, mechanical, optical and many
other types of parameters.
The measurement means, to monitor a process or a operation and using an instrument,
express the parameter, quantity or a variable in terms of meaningful numbers. Such a
measurement gives in depth knowledge of the process and the parameter and helps in
further modifications, if required. Thus the measurement provides us with a means of
expressing a natural phenomena or the various processes, in quantitative terms. The
feedback information is possible with the help of measurement techniques, which helps in
achieving goals and objectives of various engineering processes and systems.
The measurement of a given parameter or quantity is the act or result of a quantitative
comparison between a predefined standard and an unknown quantity to be measured.

For the result to be meaningful, there are two basic requirements ;-

1. The comparison standard is accurately defined and commonly accepted, and
2. The procedure and the instrument used for obtaining the comparison must be
provable.

The major problem with any measuring instrument is the error. Hence, it is necessary
to select the appropriate measuring instrument and measurement procedure which
minimises the error. The measuring instrument should not affect the quantity to be
measured.
 Key Point: The measuring instrument may be defined as a device for determining the value
or magnitude of a quantity or variable.

An electronic instrument is the one which is based on electronic or electrical principles

for its measurement function. The measurement of any electronic or electrical quantity or
variable is termed as an electronic measurement.
Quantity
to be
measured
1.1.1 Advantages of Electronic Measurement
(Measurand)
The advantages of an electronic measurement are
1. Most of the quantities can be converted by transducers into the electrical or
electronic signals.
2. An electrical or electronic signal can be amplified, filtered, multiplexed, sampled
and measured.
3. The measurement can easily be obtained in or converted into digital form for
automatic analysis and recording.
4 The measured signals can be transmitted over long distances with the help of
cables or radio links, without any loss of information.
5. Many measurements can be carried either simultaneously or in rapid succession.
6. Electronic circuits can detect and amplify very weak signals and can measure the
events of very short duration as well.
7. Electronic measurement makes possible to build analog and digital signals. The
digital signals are very much required in computers. The modern development in
science and technology are totally based on computers.
8. Higher sensitivity, low power consumption and a higher degree of reliability are
the important features of electronic instruments and measurements.
But, for any measurement, a well defined set of standards and calibration units is
essential. This chapter provides an introduction to different types of errors in
measurement, the characteristics of an instrument and different calibration standards.
Key Point : The physical, chemical, electrical quantity, property, process, variable or a
condition to be measured is referred as measurand.

The various physical measurands are pressure, level, force, strain, velocity, etc. while
the important electrical measurands are voltage, current, power and frequency.

Any instrument or a measuring system can be described in general with the help of a
block diagram. While describing the general form of a measuring system, it is not
necessary to go into the details of the physical aspects of a specific instrument. The block
diagram indicates the necessary elements and their functions in a general measuring
system. The entire operation of an instrument can be studied interms of these functional
elements. The Fig. 1.1 shows the block diagram showing the functional elements of an
instrument.
 pci~§t9rage.·
.~Q~R!ayp~9K
cal principles ••·•·
•••
•··
••
•..•.
~r~lJl~nf..••.
·•·..
I quantity or
Primary Variable Variable Data Data
Quantit
sensing conversion manipulation transmission presentation
to be
element element element element element Observer
measured
(Measurand) '------~---_.-/
Data conditioning elements

The various elements can be grouped as,

1. Pnmary sensing element
2. Data conditioning elements
3. Data presentation element
Each element is made up of number of distinct components which perform a particular
function in the measurement procedure. In the block diagram, the function of each element
is important rather than the construction of the element.

1.2.1Primary Sensing Element

signals. The
velopment in An element of an instrument which makes first, the contact with the quantity to be
measured is called primary sensing element. Thus first detection of the measurand is done
by the primary sensing element. In ammeter, coil carrying current to be measured is a
primary sensing element. In most of the cases, a transducer follows primary sensing
element which converts the measurand into a corresponding electrical signal.
Key Point: In general, a transducer converts a physical quantity from one form to other but
illcase of electrical and electronic measurements, it converts physical quantity to be measured
into an analogous electrical signal.

The output of the primary sensing element is in electrical form such as voltage,
frequency or any other electrical parameter. Such an output may not be suitable for the
actual measurement system. For example if the measurement system is digital then the
analog signal obtained from the primary sensing element is not suitable for the digital
the help of a system. Thus analog to digital converter is required which is nothing but variable
m, it is not conversion element.
nt. The block Key Point: Some instruments do not need variable conversion element while some need more
al measuring than one.
se functional
ments of an The original information about the measurand must be retained as it is while doing
such conversion.
 h to drive the next
serving the original
Current Moving
coil -
Magnets and Force
other
components
Mechanical
linkages
Pointer
and
scale
-
Primary Data Data Data
sensing conditioning transmission presentation
the signal.

f the input, at i~s

sary that variable
t. It may be before Selecting the proper instrument for a particular type of measurand needs the
knowledge of the performance characteristics of an instrument. Let us discuss the types of
the previous stage
performance characteristics of an instrument.
ed as the variable

vanted signals like

s like modulation, The performance characteristics of an instrument are mainly divided in two categories:
form from highly i) Static characteristics and ii) Dynamic characteristics.
clition to variable
Some applications involve measurands which are either constant or varying very
one in the second
slowly with time.
ing elements.
Key Point: The set of criteria defined for the instruments, which are used to measure •......•
the
/la11titicswhich are slowly varying with time or mostly constant, i.e. @ /lot vary with time, is
ca//('(/ static c aracteristic~.
ssary to transmit
sion element. The While, when the quantity under measurement changes rapidly with time, it is
iate stage of an necessary to study the dynamic relations existing between. input and output. These
relations are generally expressed with the help of differential equations.
Key Point: The set of criteria defined based on such dynamic differen tial equatio/ls is called
dynamic characteristics.
g, controlling or The various performance characteristics are obtained in one form or another by a
e information in proc(',-;scalled calibration.
function is done
I display devices 1.3.1 Calibration
analysis purpose Key Point: Calibration is the process of making an adjustment or marking a scale so that the
tation elements. readings of an instrument agree with the accepted and the certified standard.
microcontrollers
may be called In other words, it is the procedure for determining the correct values of measurand by
comparison with the measured or standard ones. The particular instrument is compared
with either a primary standard, secondary standard with higher accuracy or an instrument
t or voltage as
with known accuracy.
agnets and coil
ceoThiSforce is The calibration offers a guarantee to the device or instrument that it is operating with
ta transmission required accuracy, under the stipulated environmental conditions. It creates the confidence
of using the properly calibrated instrument, in user's mind. The periodic calibration of an
instrument is very much necessary.
 The calibration procedure involves the steps like visual inspection for various defects,
installation according to the specifications, zero adjustment etc.
The calibration characteristics can be determined by applying known values of
quantities to be measured and recording the corresponding output of the instrument. Such
output values are then compared with the input, to determine the error. Such a record
obtained from calibration is called calibration record. It is generally recorded in the tabular
form. If it is represented in the graphical form, it is called calibration curve. Such a
calibration record or calibration curve is useful to obtain the performance characteristics of
an instrument. The performance of the instrument is not guaranteed by the calibration. It
only mdicates whether the performance of the instrument is meeting the accuracy and
range specification or not. If the device has been repaired, aged, adjusted or modified, then
recalibration is carried out.

As mentioned earlier, the static characteristics are defined for the instruments which
measure the quantities which do not vary with time. The various static characteristics are
accuracy, precision, resolution, error, sensitivity, threshold, reproducibility, zero drift,
stabili ty and linearity.

1.4.1 Accuracy
It is the degree of closeness with which the instrument reading approaches the true
value of the quantity to be measured. It denotes the extent to which we approach the
actual value of the quantity. It indicates the ability of instrument to indicate the true value
of the quantity. The accuracy can be expressed in the following ways.
1) Accuracy as 'Percentage of Full Scale Reading' : In case of instruments having
uniform scale, the accuracy can be expressed as percentage of full scale reading.
For example, the accuracy of an instrument having full scale reading of 50 units
may be expressed as ± 0.1% of full scale reading. From this accuracy indication,
practically accuracy is expressed in terms of limits of error. So for the accuracy
limits specified above, there will be ± 0.05 units error in any measurement. So for
a reading of 50 units, there will be error of ± 0.05 units i.e. ± 0.1 % while for a
reading of 25 units, there will be error of ± 0.05 units in the reading i.e. ± 0.2%.
Thus as reading decreases, error in measurement is ± 0.05 units but net percentage
error is more. Hence, specification of accuracy in this manner is highly misleading.
2) Accuracy as 'Percentage of True Value' : This is the best method of specifying the
accuracy. It is to be specified in terms of the true value of quantity being
measured. For example, it can be specified as ± 0.1% of true value. This indicates
that in such cases, as readings get smaller, error also gets reduced. Hence accuracy
of the instrument is better than the instrument for which it is specified as percent
of full scale reading.
 3) Accuracy as 'Percentage of Scale Span' : For an instrument, if am,,, is the
maximum point for which scale is calibrated, i.e. full scale reading and a 111111 IS the
lowest reading on scale. Then (am<1X - amin) is called scale span or span of the
own values of
instrlJment. Accuracy of the instrument can be specified a5 percent of such scale
strument. Such
span. Thus for an instrument having range from 25 units to 225 units, it can be
Such a record
specified as ± 0.2 % of the span i.e. ± [(0.2/100) x (225 - 25)] which is ± 04 units
d in the tabular
error in any measurement.
curve. Such a
4) Point Accuracy: Such an accuracy is specified at only one particular point of scale.
haracteristics of
It does not give any information about the accuracy at any other POll1t on the
e calibra tion. It
scale. The general accuracy of an instrument cannot be specified, in this manner.
accuracy and
But the general accuracy can be specified by providing a table of the pOint
modified, then
accuracy values calculated at various points throughout the entire range of the
instrument.
Thus the accuracy, in whatever way it may be specified, gives the equipment accuracy
with a particular set up and other conditions and does not include any personal accuracy.
ments which
racteristics are
ty, zero drift, It is the measure of consistency or repeatability of measurements.
Key Point: It denotes the closeness with which individual measurements are departed or
distriuuted about the average of number of measured values.

aches the true Let us see the basic difference between accuracy and precision. Consider an instrument

approach the on which, readings upto 10100th of unit can be measured. But the instrument has large zero
the true value ]
adjustment error. Now every time reading is taken, it can be taken down upto '1000th of

unit. So as the readings agree with each other, we say that the instrument is highly
precise. But, though the readings are precise upto 10100th of unit, the readings are

inaccurate due to large zero adjustment error. Every reading will be inaccurate, due to
such error. Thus a precise instrument may not be accurate.
Thus the precision means sharply or clearly defined and the readings agree among
themselves. But there is no guarantee that readings are accurate.
An instrument having zero error, if calibrated properly, can give accurate readings but
in that case still, the readings can be obtained down upto l~OOth of unit only. Thus

accuracy can be improved by calibration but not the precision of the instrument.
Consider another example. A known weight of 100 grams is measured by an
instrument. Five times, the weight has been recorded. The readings obtained are 103, 104,
105, 103, 105. The average indicated value is 104 grams. Hence the maximum deviation
from the average reading is ±1 gram in 100 grams actual value. Thus, the scale of the
instrument can be calibrated to read ±1 gram. But what about the accuracy? The readings
are not accurate The accuracy of the instrument is only 005 - 100/100)0/" i.e. 5°/" Thus
there is a precision of ± 1% but the accuracy isonly 5%.
Key Point: This confirms the fact that high degree of precision does not guarantee the
accuracy. It is the accurate calibration that makes the accurate measurement possible.

The precision is composed of two characteristics:

• Conformity and

• Number of significant figures.

1.4.2.1 Conformity
Consider a resistor having true value as 2385692 0, which is being measured by an
ohmmeter. Now, the meter is consistently measuring the true value of the resistor. But the
reader, can read consistently, a value as 2.4 MD due to nonavailability of proper scale. The
value 2.4 MO is estimated by the reader from the available scale. There are no deviations
from the observed value. The error created due to the limitation of the scale reading is a
precision error.
The example illustrates that the conformity is a necessary, but not sufficient condition
for precision. Similarly, precision is necessary but not the sufficient condition for accuracy.
Key Point: An aCCllrllte instrument should be precise but a precise instrument may not be
accurate.

1.4.2.2 Significant Figures

The precision of the measurement is obtained from the number of significant figures,
in which the reading is expressed. The significant figures convey the actual information
about the magnitude and the measurement precision of the quantity.

For example, a resistance of 110 0, specified by an instrument may be closer to 109 Q

or 111 O. Thus there are 3 significant figures. While if it is specified as 110.0 0 then it may
be closer to 110.1 Q or 109.9 O. Thus there are now 4 significant figures.
Key Point: Thus more the sigllificant figures, the greater is the precision of meaSltrl'lIlellt.
The most irr
Number of times, the large numbers with zeros before a decimal point are used to
generally express
approximclte populations or the amounts of money. For example, the price of a vehicle is
Key Point:
reported as Rs. 450,000. This means the true value of the vehicle lies between Rs.449,999
quantity to be
and Rs. -i50,001. Thus, there are six significant figures. But what is the meaning of the
reported price is, it is closer to Rs. 450,000 rather than Rs. 440,000 or Rs. 460,000. In this Mathematica
C<lse, there <Ire only two significant figures. To avoid this confusion, the large numbers are
expressed in a scientific notation using the powers of ten. For example, the price of
Rs.450.000 must be expressed as 4.5 x 10S. Thus now, there are only two significant
figures. The uncertainty due to the zeros to the left of the decimal point is usually resolved
by such scientific notation.
 110tguarantee the
t possible.

P Precision
Xn Value of nth meilsurement
X 11 A verage of the set of measured villues

}l'.
measured by an
e resistor. But the
proper scale. The
are no deviations
scale reading is a
Example 1.1 : The table shows the set of 5 meaSl/rellll.!llts recorded 111 a laboratory.
CtJlClIltJtethe precision of the 3rd I1ll.!aSUrel1lellt.
Measurement Number Value of Measurement
1 49
2 51

3 52
ufficient condition 4 50
ion for accuracy. c::
'" 49
lIInl'1ltmay not be
Solution: The average value for the set of me?surements is,

Sum of the readings 251 ~

------- =- = -:>02
Number of readings 5 .
ignificant figures,
ctua] informa tion

e closer to 109 n P = 1_IX -=-XX n n

1= 1_152 ~- 50.21 = 0.964
-:>0.2
i.e. 96.4 %
n

.0 Q then it may This is the precision of the 3 rd measurement.

The most important static characteristics of an instrument is its accuracy, which is

oint are used to
generally expressed in terms of the error calJed static error.
e of a vehicle is
een Rs. 449,999 Key Point: The algfbmic difference between the indicated value {{nd the true va! ue of the
meaning of the quantity to be measured is called an error.
.460,000. In this Mathematically it can be expressed as,
rge numbers are
Ie, the price of e = At - Am J
two significant
usually resolved
 In this expression, the error denoted as e is also called absolute error. The absolute
error does not indicate precisely the accuracy of the measurements. For example, absolute
error of ± 1 V is negligible when the voltage to be measured is of the order of 1000 V but
the same error of ± 1 V becomes significant when the voltage under measurement is 5 V or
so. Hence, generally instead of specifying absolute error, the relative or percentage error is
specified.
Milthematically, the relative error can be expressed as,

Absolute Error True value - t¥[easured value

True value True value

A I - Am X 100
At

A 1 - e, = 1 _I A I ;. t
A
m I
I
where A Relative accuracy

and a A x 100 %
I
where a Percentage aCCl.1racy

Error as a percentage of full scale reading = At - A m x 100

f.s.d.

llJ. Example 1.2: The expected value of the voltage to be measured is 150 V. However, the
JIIeasurement gives a value of 149 V. Calculate (i) absolute error (ii) percentage error
(iii) relative accuracy (iv) percentage accuracy and (v) error expressed as percentage of full
scale reading, if the scale range is 0-200 V.

Solution : The expected value means true value,

 If the calibration curve is not linear as shown in the Fig. 1.3 (b), then the sensitivity
varies with the input.
The sensitivity is always expressed by the manufacturers as the ratio of the magnitude
of quantity being measured to the magnitude of the response. Actually, this definition is
the reciprocal of the sensitivity is called inverse sensitivity or deflection factor. But
m'lllufacturers call this inverse sensitivity as a sensitivity.

I =~
Sensitivity 6. q 0

The units of the sensitivity are millimeter per micro-ampere, millimeter per ohm,
counts per volt, etc. while the units of a deflection factor are micro-ampere per millimeter,
ohm per millimeter, volts per count, etc.
1.4.6 1i
The sensitivity of the instrument should be as high as possible and to achieve this the
If the
r,mge of an instrument should not greatly exceed the value to be measured.

m. Example 1.3: A particular ammeter requires a change of 2 A in its coil to produce a

change ill deflection of the pointer by 5 mm. Determine its sensitivity and deflection factor.
until son:
called thl
Key
[s tire
Solution: The input is current while output is deflection.

Change in output
Sensitivity = -------
Change in input
The
accordin
= 5 mm = 2.5 mm / A
2A charade]
output i!
The
Key
I
- = 0.4 A / mm
2.5 The
(output)
a percen
It is the smallest increment of quantity being measured which can be detected with The
ccrtainil)' by an instrument.
Key Point: Thus, the resolution means the smallest measurable input change.

50 if a nonzero input quantity is slowly increased, output reading will not increase
until some minimum change in the input takes place. This minimum change which causes
the change in the output is called resolution. The resolution of an instrument is also
referred to as discrimination of the instrument. The resolution can affect the accuracy of
the measurement.
 2. Time Delay : In this case, response begins after some time called dead time, after
the application of input. Such a delay shifts the response along time axis and
hence causes the dynamic error.

1.5.4 Dynamic Error

It is the difference between the true value of the variable to be measured, changing
with time and the value indicated by the measurement system, assuming zero static error.
The Fig. 1.13 shows the dead time, i.e. time delay and the dynamic error.

Input
and
Output

Instrument reading
I

:
T
Dynamic error
I t

-:
0: Dead
time
---
I
I

The static error is defined earlier as the difference between the true value of the
variable and the value indicated by the instrument. The static error may arise due to
number of reasons. The sta tic errors are classified as :
1) Gross errors
2) Systematic errors
3) Random errors
Let us discuss these types of errors.

The gross errors mainly occur due to carelessness or lack of experience of a human
being. These cover human mistakes in readings, recordings and calculating results. These
errors also occur due to incorrect adjustments of instruments. These errors cannot be
treated mathematically. These errors are also called personal errors. Some gross errors are
easily detected while others are very difficult to detect.
The complete elimination of gross errors is not possible but one can minimise them by
the fWowing ways:
1. Taking great care while taking the reading, recording the reading and calculating
the result.
 2. Without depending on only one reading. At least three or even more readings
must be taken and preferably by different persons. The readings must be taken
preferably under the conditions in which the instruments are switched on and off.

1.6.2 Systematic Errors

The systematic errors are mainly resulting due to the shortcomings of the instrument
and the characteristics of the material used in the instrument, such as defective or worn
parts, ageing effects, environmental effects, etc.
A constant uniform deviation of the operation of an instrument is known as a
systematic error.
There are three types of systematic errors as
1) Instrumental errors 2) Environmental errors 3) Observational errors

1.6.2.1 Instrumental Errors

These errors can be mainly due to the following three reasons.
a) Shortcomings of Instruments : These are because of the mechanical structure of
the instruments. For example, friction in the bearings of various moving parts,
irregular spring tensions, reduction in tension due to improper handling,
hysteresis, gear backlash, stretching of spring, variation in air gap, etc. These errors
can be avoided by the following methods.
1) Selecting a proper instrument and planning the proper procedure for the
measurement.

2) Recognizing the effect of such errors and applying the proper correction
factors.
3) Calibrating the instrument carefully against a standard.
b) Misuse of Instruments : A good instrument if used in abnormal way gives
misleading results. Poor initial adjustments, improper zero setting, using leads of
high resistance etc. are the examples of misusing a good instrument. Such things
do not cause permanent damage to the instruments but definitely cause serious
errors.
c) Loading Effects : Loading effect due to improper way of using the instrument
cause the serious errors. The best example of such loading effect error is
connecting a well calibrated voltmeter across the two points of high resistance
circuit. The same voltmeter connected in a low resistance circuit gives accurate
reading. Thus, the errors due to the loading effect can be avoided by using an
instrument intelligently and correctly.

These errors are due to the conditions external to the measuring instrument. The
various factors resulting these environmental errors are temperature changes, pressure
 changes, thermal e.m.L stray capacitance, cross capacitance, effect of external fields, agemg
of equipment and frequency sensitivity of an instrument.
The various methods which can be used to reduce these errors are
T. Using the proper correction factors and using the information supplied by the
manufacturer of the instrument.
') Using the arrangements which will keep the surrounding conditions constant. This
includes the use of air conditioning, temperature control enclosures etc.
1 Reducing the effect of dust, humidity on the components by hermetically sealing
the cOl1,ponents in the instruments.
4. The effects of external fields can be minimised by using the magnetic or
electrostatic shields or screens.
5. Using the equipment which is immune to such environmental effects. For example,
in environment having lot of temperature variations, use of an instrument in which
resistance material having a very low resistance temperature coefficient is
appropriate.

These are the errors introduced by the observer. There are many sources of
observational errors such as parallax error while reading a meter, wrong scale selection,
the habits of individual observers etc.
To eliminate such observational errors, one should use the instruments with mirrors,
knife edged pointers, etc. Nowadays, the instruments with digital display are available
which can largely eliminate such observational errors.
The systematic errors can be subdivided as static and dynamic errors. The static errors
are caused by the limitations of the measuring device while the dynamic errors are caused
by the instrument not responding fast enough to follow the changes in the variable to be
measured.

Some errors still result, though the systematic and instrumental errors are reduced or
atleast accounted for. The causes of such errors are unknown and hence, the errors are
called random errors. These errors cannot be determined in the ordinary process of taking
the measurements.
These errors are generally due to the accumulation of large number of the small
effects. These errors are generally small. Hence, these errors are of real concern only when
the high degree of accuracy is required.
The random errors follow the laws of probability and hence, these errors can be
analyzed statically and tre~afe-dmathematically. These- eno:rs cannot be _CD.rre~tedby any
method of calibration or other known method of control as the causes of such errors are
unkn.own.
 The only way to reduce these errors is by increasing the number of observations and
using the statistical methods to obtain the best approximation of the reading.

When the error is specified interms of an absolute quantity and not as a percentage,
then it is called an absolute error.
Thus the voltage of 10 ± 0.5 V indicated ± 0.5 V as an absolute error.
When the error is expressed as a percentage or as a fraction of the total quantity to be
measured, then it is called relative error.

Thus if the resistance is specified as 100 D ± 5 % then ± 5 % or ± (;0) of 100 D is the

relative error.
Key Point : Many a times absolute error is specified as accuracy while the relative error IS

termed as tolerance.

Generally the relative error in case of

resistances is specified as percentage
tolerances.

Another method of expressing error is by

specifying it as parts per million (ppm),
.6.T relative to the total quantity. So it is a
Slope = .6.R in ppm/oC
.6.T
relative error specification .
Generally change in resistance with
temperature is indicated in ppm. The Fig. 1.14
°C shows the variation in resistance with
Temperature
temperature. Thus if a resistance of 100 kD. has
a temperature coefficient of 50 ppm/"C means
50 parts per millionth per degree celcius. Thus
one millionth of 100 kD. is 0.1 D and 50 such parts means 5 D. Th1-1s 1"<:::change in
temperature causes change of 5 D in 100 kD. resistor.

))). Example 1.5 : A resistance manufacturer has stated the range of a resistance betwec1/
2.61 kn to 2.79 kQ and specifies it as 2.7 kQ resistor. What is thc tolerance of rcsistor 7 Jf
thc resistance valllcs are specified at 27 DC and resistors havc temperature coc!ficicllt of
+ 100 ppm 1°C, find the maximum value of one of these resistors at 82 uc.
Solution : The range is 2.61 kD. to 2.79 kD..

Absolute error
------- x 100
Quantity specified
 ±0.09
--- x 100 = ± 3.33 %
2.7

2.7 + 0.09 = 2.79 kD.

100
100 ppm = -' --x 2.79x 103
1 x 106
.(!ill./ 6T)

Change in temperature 82 - 27 = 55 "C ... (6T)

Total change in resistance = 55 x 0.279 = 15.345 n ... (!ill.)

The manufacturers specify the accuracy of the instruments within a certain percentage
of full scale reading. The components like the resistor, inductor, capacitor are guaranteed
to be within a certain percentage of rated value. This percentage indicates the deviations
from the nominal or specified value of the particular quantity. These deviations from the
specified value are called Limiting Errors. These are also called Guarantee Errors.
For example, the manufacturer of a certain instrument may specify that the instrument
is accurate within ± 1% of full scale deflection. This means that a full scale reading is
guaranteed to be within ± 1% of a perfectly accurate reading. But for a reading less than
full scale, the limiting error increases.
Ano.ther example is say a resistor is specified by the manufacturer as 4.7 kD. with a
tolerance of ± 5% then the actual value of the resistance is guaranteed to be within the
limits.
R 4.7 kD. ± (5 % of 4.7 kD. ) = 4.7 kD. ± 0.235 kn
4.935 kD. and 4.465 kD.

Thus the actual value with the limiting error can be expressed mathematically as,

I Aa = As ± 8AI
Aa = Actual value

As Specified or rated value

1.8.1 Relative Limiting Error
This is also called fractional error. It is the ratio of the error to the specified
magnitude of a quantity.

Thus Ie" SA
As I

As ± e As

I Aa = As [ 1 ± ell
The percentage relative limiting error is expressed as

1% e = e x 100 I

Actual value (A a ) - Specified value (A s )

Specified value (As)

m. Example 1.6 : A 0-50 V voltmeter is specified to be accurate within

Ca/cII/ate the Limiting error when the instrument reading is 15 V.
± 1 % of fIlII scaLe.

Solution : The limiting error at full scale is,

I
± 100 x 50

For a reading of 15 V, it is

% e = 0.5 x 100 = 3.33 %

15
 When the two quantities are combined, each having limiting error, then it i" nl'cessan'
to calculate the overall limiting errOL Let us consider the various combinationo, of two
qll<1lltities,1lld methods to obtain the corresponding limiting errOL

Let ill '1lld il2 be the two quantities which Me to be added to obtain the result il" A I'
Cono,ider the reliltive increment of the function which the riltio of change in function to
the \'illuc of the function,

c1(<11 +<12)
AT

Let 0 a I be the limiting error of a,

,1Ild is a2 be the limiting error of a2

Hence, the corresponding relative limiting errors are

iSA
as e =-
As

is AT
eT =
AT

Lli 8 Ll I 3, 832
eT --+-- . '"
from (I)
AT 31 AT 32

a?
eT + [3- , e, +-- , e2 ]
- AT' AT

Key Point: Thus, the resultant (fatal) limiting error is slim of the products obtained by
lIIultiplying the individllal /ill/iting error by the ratio of each term to the resliitant flincliaiL

a1 - a2
d(al - a2) clal da2
-----
AT AT AT
 ow, if ± 8211 and ± 8212 are the errors in 211 and 212 respectively, then maximum
possible error will result when the- signs of 8211 and 8212 are opposite to each other. For
same signs of 8 211 and 8 212 the error will be very small. Hence, conside'ring worst possible 4.
discrepancy i.e. when 8 211 is positive, 8212 is negative and vice-versa.

But, as 8211 and 8212 are of opposite sign, the resultant sign can be taken common to
express the result as,

If a, and 212 are almost same, then as AT is 211 - 212' AT « 211 and AT « 212 also, then
the relative error eT in AT would be very large.
If there is difference of ,more than two quantities,then

8a 0
e2 = ---I
a2

Differentiating with respect to AT'

1 1 dal 1 da2
- = -.--+-.--
AT al dAT a2 dAT

Multiplying both sides by dAT,

dAT dal dao
-- + ---
AT al a2

Thus, if 821, and 8212 are the limiting errors of 211 and 212' then,

8AT 8al 8a7

-- + ---
AT a, a2
 leT = ± (e] + e2) I
Key Point: Thus, the relative limiting error of the product is eqllal to the SIIIII of ti,e relative
lilliiting errors.

log Ar = log [:~ ] = log 0] - log a2

= ~_ d32
3I 32

oat _ 032

aI 32

But, clgain consider the worst case i.e. oa] and oa2 are of opposite signs. Hence, laking
the common sign outside, the result can be expressed as

leT = ± (e] + e2) I

The result is same as that obtained for the product.

If there is product or division of more than two quantities, then,

AT a] a2 a3
31
or AT
a2 a 3 .

t)J"
AT
a I a 2 33 .

then ± [ e] + ez +
leT e3 + . . ]
I
5. Power of a factor

Let AT (a1)n

log Ar Jog faIt = n log a]

I 1 d 3I
n.-.--
AT al d AT
 n
-. d al
al

where er is relative limiting error in result AT while e, is relative limiting error in a,.
If the result is the product of different powers of two quantities i.e.

AT = (a,t· (a2)111

Then ,1S it is a product of two quantities, each having some power, the resultant
limiting error can be expressed as,

leT = ± [ n el + m e2 ]

)))•• Example 1.7: The r.m.s. current passing through II resistor of 120 ± 0.5 ohms is
2 ± 0.U2 A. Calculate the limiting error in the vallie of power dissipation.

8 a2 Limiting error in resistor = 0.5

el = ~ = 0.02 = 0.01
AI 2

The current term in power appears as r2 so it is second power of l.

[fence, the contribution by r2 to the resultant error is nel

while the limiting error due to resistance is e2.

As power is the product of r2 and R, the resultant error is the sum of the contributions
1
by f- c1nd R.