CHAPTER 2

MEASUREMENT AND ERROR

No electronic component or instrument is perfectly accurate. All have some error or

inaccuracy. It is important to understand how these errors are specified and how they
combine to create even greater errors in measurement systems. Although it is pos-
sible that in some cases errors might almost completely cancel each other out. The
worst case combination of errors must always be assumed. Apart from equipment
errors, some operator or observer error is inevitable. Also even when equipment er-
rors are very small, the system of using the instruments can introduce a systematic
error. Errors of unexplainable origin are classified as random errors. Where accuracy
is extremely important, some errors can be minimized by taking many readings of
each instrument and determining mean values.

2.1 Accuracy and Precision

Accuracy refers to the degree of closeness or conformity to the true value of the
quantity under measurement. When a voltmeter with an error of ±1% indicates ex-
actly 100 V, the true level of the measured voltage is somewhere between 99 V and
101 V. Thus, the measurement accuracy of ±1% defines how close the measurement
Measurements and Instrumentation, First Edition. 9
By Osama A. Alkishriwo Copyright c 2014 John Wiley & Sons, Inc.
10 MEASUREMENT AND ERROR

is to the actual measured quantity.

The precision refers to the degree of agreement within a group of measurements (i.e.
given a fixed value of variable, Precision is measure of the degree to which succes-
sive measurements differ from one another). Consider, for example, two voltmeters
one is a digital voltmeter and the other is an analog voltmeter. For the 8.135 V
quantity indicated, the last (right–side) numeral refers to millivolts. If the measured
quantity increases or decreases by 1 m V, the reading becomes 8.136 V or 8.134 V,
respectively. Therefore, the voltage is measured with a precision of 1 mV. For the
analog voltmeter, the pointer position can be read to within (perhaps) one–fourth of
the smallest scale division. If the smallest scale division represents 0.2 V (on the 10
V range), one–fourth of the scale division is 50 mV. So, 50 mV is the measurement
precision of the analog instrument.

2.2 Resolution

The measurement precision of an instrument defines the smallest change in measured

quantity that can be observed. This (smallest observable change) is the resolution of
the instrument. In the case of the 10 V analog voltmeter scale that can be read to a
precision of 50 mV where 50 mV is the smallest voltage change that can be observed.
Thus, the measurement resolution is 50 mV.

2.3 Significant Figures

An indication of the precision of the measurement is obtained from the number of

significant figures in which the result is expressed. The more significant figures,
the greater the precision of measurement. In the case of a resistance value stated as
47.3 Ω, the actual value may not be exactly 47.3 Ω, but it is assumed to be closer
to 47.3 Ω than it is to either 47.2 Ω or 47.4 Ω. The three significant figures show
that measurement precision is 0.1 Ω. If the quantity was 47.3 kΩ, the measurement
precision would be 0.1 kΩ, or 100 Ω. If 47.3 Ω is rewritten with two significant
figures, it becomes 47 Ω, because 47.3 Ω is clearly closer to 47 Ω than it is to 48 Ω.

Example 2-1
Two resistors. R1 and R2 are connected in series. Individual resistance mea-
surements using a digital multimeter give R1 = 18.7 Ω and R2 = 3.624 Ω.
Calculate the total resistance to the appropriate number of significant figures.

Solution
R1 = 18.7 Ω (three significant figures)
R2 = 3.624 Ω (four significant figures)
RT =R1 + R2 = 22.324 Ω (five significant figures)= 22.3 Ω
 MEASUREMENT ERROR COMBINATIONS 11

2.4 Measurement Error Combinations

When a quantity is calculated from measurements made on two (or more) instru-
ments, it must be assumed that the errors due to instrument inaccuracy combine in
the worst possible way. The resulting error is then larger than the error in any one
instrument.

1. Sum of Quantities
When a quantity is determined as the sum of two measurements, the total error
is the sum of the absolute errors in each measurement.

E = (V1 ± ∆V1 ) + (V2 ± ∆V2 )

= (V1 + V2 ) ± (∆V1 + ∆V2 ) (2.1)

Example 2-2
Calculate the maximum percentage error in the sum of two voltage mea-
surements when V1 = 100 V ± 1% and V2 = 80 V ± 5%.
Solution

V1 = 100 V ± 1%
= 100 V ± 1 V
V2 = 80 V ± 5%
= 80 V ± 4 V
E = V1 + V2
= (100 V ± 1 V ) + (80 V ± 4 V )
= 180 V ± (1 V + 4 V )
= 180 V ± 5 V
= 180 V ± 2.8%

2. Difference of Quantities
There are situations in which a potential difference is determined as the differ-
ence between two measured voltages. Here again, the errors are additive:

E = (V1 ± ∆V1 ) − (V2 ± ∆V2 )

= (V1 − V2 ) ± (∆V1 + ∆V2 ) (2.2)

Example 2-3
Calculate the maximum percentage error in the difference of two measured
voltages when V1 = 100 V ± 1% and V2 = 80 V ± 5%.
12 MEASUREMENT AND ERROR

Solution

V1 = 100 V ± 1%
= 100 V ± 1 V
V2 = 80 V ± 5%
= 80 V ± 4 V
E = V1 − V2
= (100 V ± 1 V ) − (80 V ± 4 V )
= 20 V ± (1 V + 4 V )
= 20 V ± 5 V
= 20 V ± 25%
Example 2−3 demonstrates that the percentage error in the difference of two
quantities can be very large. If the difference was smaller, the percentage
error would be even larger. Obviously, measurement systems involving the
difference of two quantities should be avoided.

3. Product of Quantities
When a calculated quantity is the product of two or more quantities, the per-
centage error is the sum of the percentage errors in each quantity.
P = EI
= (E ± ∆E) (I ± ∆I)
= E I ± E ∆I ± I ∆E ± ∆E ∆I
Since ∆E ∆I is very small
P ' E I ± (E ∆I + I ∆E)
E ∆I + I ∆E
Percentage error = × 100%
 EI 
∆I ∆E
= + × 100%
I E
% error in P = (% error in I) + (% error in E) (2.3)

4. Quotient of Quantities
Here again it can be shown that the percentage error is the sum of the percentage
errors in each quantity.
% error in E/I = (% error in I) + (% error in E) (2.4)

5. Quantity Raised to a Power

When a quantity A is raised to a power B, the percentage error in AB can be
shown to be
% error in AB = B (% error in A) (2.5)
 TYPES OF ERROR 13

Example 2-4
An 820 Ω resistance with an accuracy of ±10% carries a current of 10 mA.
The current was measured by an analog ammeter on a 25 mA range with an
accuracy of ±2% of full scale. Calculate the power dissipated in the resistor,
and determine the accuracy of the result.

Solution

P = I2 R
P = (10 mA)2 × 820Ω
= 82 mw
error in R = ±10%
error in I = ±2% of 25 mA
= ±0.5 mA
±0.5 mA
= × 100%
10 mA
= ±5%
2
% error in I = 2(±5%)
= ±10%
% error in P = (% error in I 2 ) + (% error in R)
= ±(10% + 10%)
= ±20%

2.5 Types of Error

No measurement can be made with perfect accuracy but it is important to find out
what accuracy actually is and how different errors have entered into the measure-
ment. A study of errors is a first step in finding ways to reduce them. Errors may
arise from different sources and are usually classified as under:

2.5.1 Gross Errors

This class of errors mainly covers human mistakes in reading instruments and record-
ing and calculating measurement results. The responsibility of the mistake normally
lies with the experimenter. For example, he may read 25.8 C o and record 28.5 C o .
But as long as human beings are involved, some gross errors will definitely be com-
mitted. Although complete elimination of gross errors is probably impossible, one
should try to anticipate and correct them. Some gross errors are easily detected while
others may be very difficult to detect. Gross errors may be of any amount and there-
14 MEASUREMENT AND ERROR

fore their mathematical analysis is impossible. However, they can be avoided by

adopting two means. They are:
1. Great care should be taken in reading and recording the data.
2. Two, three or even more readings should be taken for the quantity under mea-
surement.

2.5.2 Systematic Errors

These types of errors are divided into three categories:
1. Instrumental errors: These errors arise due to three main reasons:
Due to inherent shortcomings in the instrument,
Due to misuse of the instruments,
Due to loading effects of instruments.
2. Environmental errors: These errors are due to conditions external to the mea-
suring device including conditions in the area surrounding the instrument. These
may be.effects of temperature pressure, humidity, dust, vibrations or of external
magnetic or electrostatic fields. The corrective measures employed to eliminate
or to reduce these undesirable effects are:
Arrangements are made to keep the conditions as nearly as constant as pos-
sible. For example, temperature can be kept constant by keeping the equip-
ment in a temperature controlled enclosure.
Using equipment which is immune to these effects. For example, varia-
tions in resistance with temperature can be minimized by using resistance
materials which have a very low resistance temperature coefficient.
Employing techniques which eliminate the effects of these disturbances. For
example, the effect of humidity dust etc. can be entirely eliminated by her-
metically sealing the equipment.
3. Observational errors: Observational errors are those errors introduced by the
observer. The two most common observational errors are probably the paral-
lax error introduced in reading a meter scale and the error of estimation when
obtaining a reading from a meter scale.

2.5.3 Absolute Errors and Relative Errors

If a resistor is known to have a resistance of 500 Ω with a possible error of ±50 Ω, the
±50 Ω is an absolute error. This is because 50 Ω is stated as an absolute quantity, not
as a percentage of the 500 Ω resistance. When the error is expressed as a percentage
or as a fraction of the total resistance, it becomes a relative error. Thus, the ±50 Ω is
±10%, relative to 500 Ω, or ±1/10 of 500 Ω.
 TYPES OF ERROR 15

2.5.4 Random Errors

Random errors are those that remain after the gross and systematic errors have been
substantially reduced, or at least accounted for. Random errors are generally the
accumulation of a large number of small effects and may be of real concern only in
measurements requiring a high degree of accuracy. Such errors can only be analyzed
statistically.

Example 2-5
A voltmeter having a sensitivity of 1000 Ω/V reads 100 V on its 150 V scale
when connected across an unknown resistor in series with a milliammeter. When
the milliammeter reads 5 mA, calculate

(a) apparent resistance of the unknown resistor,

(b) actual resistance of the unknown resistor,
(c) error due to the loading effect of voltmeter.

Solution

(a) Total circuit resistance:

ET 100
RT = = = 20kΩ
IT 5 × 10−3

Neglecting the resistance of milliammeter, the value of unknown

resistance is:

Rx = 20kΩ

(b) Resistance of voltmeter Rv = 1000×150 = 150×103 Ω = 150kΩ

As the voltmeter is in parallel with the unknown resistance, we
have:

Rx Rv
RT =
Rx + Rv
RT Rv 20 × 150
or unknown resistance Rx = = = 23.077kΩ
Rv − RT 150 − 20

erronous quantity-true quantity

(c) Percentage error = true quantity × 100

23.077 − 20
= × 100 = 13.33%
23.077
16 MEASUREMENT AND ERROR

2.6 Statistical Analysis

The experimental data is obtained in two forms of tests :

Multisamp1e Test: In this test, repeated measurement of a given quantity are

done using different test conditions such as employing different instruments/
different ways of measurement and by employing different observers.

Single Sample Test: A single measurement (or succession of measurements)

done under identical conditions excepting for time is known as single-sample
test.

In order to get the exact value of the quantity under measurement, tests should be
done using as many different procedures, techniques and experimenters as practica-
ble.

2.6.1 Arithmetic Mean

When a number of measurements of a quantity are made and the measurements are
not all exactly equal, the best approximation to the actual value is found by calculat-
ing the average value, or arithmetic mean, of the results. For n measured values of
x1 , x2 , x3 , · · · , xm the arithmetic mean is
Pn
x1 + x2 + x3 + · · · + xn xi
x̄ = = i=1 (2.6)
n n
where
x̄ : arithmetic mean
xi : readings or samples
n : number of readings

2.6.2 Deviation from the Mean

The difference between any one measured value and the arithmetic mean of a series
of measurements is termed the deviation. The deviations (d1 , d2 , d3 , · · · , dn ) may
be positive or negative and the algebraic sum of the deviations is always zero.

d1 = x1 − x̄ d2 = x2 − x̄ d3 = x3 − x̄ dn = xn − x̄ (2.7)

Example 2-6
A set of independent current measurements was taken by sex observers and
recorded as 12.8 mA, 12.2 mA, 12.5 mA, 13.1 mA, 12.9 mA, and 12.4 mA.
Calculate

(a) Arithmetic mean

(b) The deviations from the mean
 STATISTICAL ANALYSIS 17

Solution
(a) The arithmetic meas is:

12.8 + 12.2 + 12.5 + 13.1 + 12.9 + 12.4

x̄ = = 12.65 mA
6
(b) The deviations are:

d1 = 12.8 − 12.65 = 0.15 mA

d2 = 12.2 − 12.65 = −0.45 mA
d3 = 12.5 − 12.65 = −0.15 mA
d4 = 13.1 − 12.65 = 0.45 mA
d5 = 12.9 − 12.65 = 0.25 mA
d6 = 12.4 − 12.65 = −0.25 mA

2.6.3 Average Deviation

The average deviation is an indication of the precision of the instruments used in
making the measurements. Highly precise instruments will yield a low average de-
viation between readings. By definition, average deviation is the sum of the absolute
values of the deviations divided by the number of readings. It can be expressed as
Pn
|d1 | + |d2 | + |d3 | + · · · + |dn | |di |
D = = i=1 (2.8)
n n
Example 2-7
Calculate the average deviation for the data given in Example 2-6.

Solution

0.15 + 0.45 + 0.15 + 0.45 + 0.25 + 0.25

D = = 0.283 mA
6

2.6.4 Standard Deviation

The mean–squared value of the deviations can also be calculated by first squaring
each deviation value before determining the average. This gives a quantity known as
the variance. Taking the square root of the variance produces the root mean squared
(rms) value, also termed the standard deviation (σ).
r r Pn
2
d21 + d22 + d23 + · · · + d2n i=1 di
σ = = (2.9)
n n
18 MEASUREMENT AND ERROR

2.7 Problems

2.1 A batch of resistors that each have a nominal resistance of 330 Ω are to be
tested and classified as ±5% and ±10% components. Calculate the maximum
and minimum absolute resistance for each case.
2.2 Three of the resistors referred to in Problem 2.1 are connected in series. One
has a ±5% tolerance, and the other two are ±10%. Calculate the maximum and
minimum values of the total resistance.
2.3 A 470 Ω, ±10% resistor has a potential difference of 12 V across its termi-
nals. If the voltage is measured with an accuracy of ±6%, determine the power
dissipation in the resistor, and specify the accuracy of the result.

2.4 The output voltage from a precision 12 V power supply, monitored at intervals
over a period of time, produced the following readings: V1 = 12.001 V, V2 =
11.999 V, V3 = 11.998 V, V4 = 12.003 V, V5 = 12.002 V, V6 = 11.997 V,
V7 = 12.002 V, V8 = 12.003 V, V9 = 11.998 V, and V10 = 11.997 V. Calculate
the average voltage level, the mean deviation, and the standard deviation.

2.5 Prove that

(a) The combination error of two quotient of quantities is given by:

% error in E/I = (% error in I) + (% error in E)

(b) The combination error of quantity A raised to a power B is given by:

% error in AB = B (% error in A)