C H AP TE R

STATIC PERFORMANCE
CHARACTERISTICS AND ERRORS
IN MEASURING INSTRUMENTS
    OVERVIEW    
In this chapter we shall learn :
ä Static performance characteristics such as linearity, sensitivity, span, resolution,
threshold accuracy, precision, hysteresis and drift.
ä Sources of error in measurement process.
ä Statistical analysis of errors.

2.1 INTRODUCTION
Whenever we go to market to purchase a measuring instrument, there are various criteria
on basis of which we compare the performance of the measuring instrument manufactured
by different companies. These critera of comparison are called performance characteristics as
they are indicative of how well an instrument measures the desired input and how well it
rejects the noise (disturbance) inputs. The performance characteristics of measuring instrument
can be broadly classified as :
1. Static performance characteristics
2. Dynamic performance characteristics.
There are some instruments in which measured quantity remains constant or varies very
slowly with respect to time. Under such condition, we examine the parameters called static
characteristics. Accuracy, precision, linear sensitivity, resolution, hyteresis and drift etc. are
examples of static performance characteristics which can be employed to describe the quality
of such measuring instruments. In case, the instruments measure rapidly varying quantities,
45
 46 Principles and Practices of Measurement, Metrology and Control

it becomes essential to study dynamic relationship between input and output using the
differential equations. The criteria used to define such continuously varying relationships are
described by dynamic performance characteristics.
Let us first study static characteristics.
1. Linearity
It is one of the most desirable features of a measuring intsrument. Every manufacturer
tries to design and manufacture the instrument such that the output is a linear function of
input. If x is the input and y is the output for a measuring instrument, then
y = mx
is a linear relation between output and input where m is the slope of the straight line with
this equation. However, it is never possible to achieve complete linearity. Some deviation is
always there from the linearity, which is often specified in terms of linearity deviation on the
measuring instrument. If the input is taken as absica and output as ordinate, then the least
square method is preferably used, to fit the straight line to the data points. Linearity is the
measure of maximum deviation of any data point from this straight line.
Mathematically,
Ma xim u m Devia t ion
Percentage Linearity = ´ 100
F ull sca le rea din g
Least square method is an important numerical technique. It would be appropriate to
discuss procedure of fitting a straight line to a given data using least square method. Let y
= P0 + P1x be the equation of a straight line that has to be fitted to a given data. Let there
be some data point (x1, y1), (x2, y2), (x3, y3) ... (xn, yn). Let curve represented by equation y
= f(x) be fitted to these data points. Then the error of approximation at some point (xi, yi) will
be given by [ti – f(xi)]. By method of least squares, the sum of square of all such errors ixs
minimixed, i.e., if S = [yi – f(x1)]2 + [y2 – f(x2)]2 + .... + [yn – f(xn)]2 then by method of least
squares, a curve is fitted such that S is minimum.

d 2S d2S
To minimise S, and should be positive. Now
d P1 2 d P2 2
S = [y1 – (P0 + P1x1)]2 + [y2 – (P0 + P1x2)]2 + [y3 – (P0 + P1x3)]2 + .... + [yn – (P0 + P1x)]2
dS dS
To start with, put dP = 0 and d P = 0
0 1

dS
= 0  –2[y1 – (P0 + P1x1)] – 2[y2 – (P0 + P1x2)]
d P0
– 2[y3 – (P0 + P1x3)] + ... – 2[yn – (P0 + P1xn)] = 0
dS
= 0  –2x1[y1 – (P0 + P1x1)] – 2x2[y2 – (P0 + P1x2)]
d P1
– 2x3[y3 – (P0 + P1x3)] + ... – 2xn[yn – (P0 + P1xn)] = 0
The above two equations can be rearranged as
nP0 + P1x = y
P0x + P1x2 = xy
 Static Performance Characteristics and Errors in Measuring Instruments 47
These two equations can be used to find the value of unknown i.e, P0 and P1. The value
of x, y, x2 and xy can be obtained from the given data.
2. Static Sensitivity
It is defined as ratio of output to input signal. Mathematically, it is given by slope of the
straight line which represents input-output relationship. If the graph showing input-output
curve is as shown in figure 2.1, the static sensitivity is given by
Ch a n ge in out put Dy
Static sensitivity (K) = =
Ch a n ge in input Dx

y y
Output

Output
y y

x x

Input x Input x

Fig. 2.1.

EXAMPLE : The following output voltage (millivolts) was obtained when an input load (kgf)
was applied to a load cell. Find the percentage linearity and the sensitivity of the instrument.
Load 12 18 24 35 40 45
Voltage 1.0 1.9 2.9 4.0 4.9 6.9
Solution: Let the equation of best fit line is y = mx + C, then using least square method
y = mx + nC ...(A)
2
xy = mx + Cx ...(B)
To find x, x, x2 and xy, we make the following table
x y x2 xy y’ = mx + C y – y’
12 1.0 144 12.0 0.872 0.127
18 1.9 324 34.2 1.832 0.067
24 2.9 576 69.6 2.792 0.1075
35 4.0 1225 140.0 4.552 – 0.5524
40 4.9 1600 196.0 5.352 – 0.452
45 6.9 2025 310.5 6.152 0.748
x = 174 y = 21.6 x2 = 5894 xy = 762.3

Substituting these values in eq. (A) and (B), we get

21.6 = m(174) + 6C
762.3 = m(5894) + C(174)
On solving, we get m = 0.16 and C = – 1.047
Thus linear equation for the instrument is
y = 0.16x – 1.047
 48 Principles and Practices of Measurement, Metrology and Control

Now we extend the table by calculating y’ for each value of x which lies on the straight
line y’ = 0.16x – 1.047 and deviation from actual value (y – y).
Ma xim um devia t ion
Percentage linearity = ´ 100
F ull sca le r ea din g
0.748
= × 100 = 12.58%
6.152
Sensitivity = Slope of the straight line
Dy
Slope = m = = 0.16 mV/kgf.
Dx
3. Span
It is defined as range of value of input signal that an instrument is designed to measure.
Hence, it is the difference between the highest value and the lowest value of the input signal
that can be measured by an instrument.
4. Resolution
It is defined as smallest increment in the input signal which can be satisfactorily detected
by the measuring instrument. In other words, it is the last count of the measuring instrument.
The resolution of an instrument is said to be high if it can measure even a smallest change
in the input. For example, resolution of micrometer is 0.01 mm. Readability of the scale is
important. Resolution indicates the ease with which the values can be read from the scale.
5. Threshold
It is defined as minimum value of input signal which can be detected by the measuring
instrument. For any value of input below the threshold value, no output is generated. It is
not possible to reduce the value of resolution and threshold to zero because of inertia,
backlash, friction between moving parts.
6. Accuracy
Accuracy of a measuring instrument may be defined as degree of closeness of the output
to the true value of measured quantity. Usually, it is specified as percentage deviation or
inaccuracy in measurement from true value of measured quantity. Mathematically, it is
given by the relation
Mea su red va lu e - Tr u e va lu e
Percentage Accuracy = × 100
Tr ue va lu e
Accuracy of the measuring instrument is disturbed by various factors such as zero errors,
tolerances given to moving parts, friction, backlash, parallex effect etc. If more than one such
factors are present, the effect may be cumulative and increase the over-all inaccuracy of the
instrument.
7. Precision
Precision of a measuring instrument is defined as its ability to reproduce the same
output repeatedly for the same input. For example, if a micrometer is used to measure the
diameter of a rod a number of times and every time it produces exactly the same output, it
is said to have high precision, even though it may be having low accuracy. The difference
between accuracy and precision may be understood from example of a shooter who has to hit
at centre of the target plate. The target plate is shown in figure 2.2, with different cases of
accuracy and precision. The black dots show the position where the bullets hit the plate.
 Static Performance Characteristics and Errors in Measuring Instruments 49

High accuracy Low accuracy Low accuracy

high precision high precision Low precision
A B C
Fig. 2.2.

From figure 2.2(b), it is apparent that even if accuracy is low, the precision can be high
as the bullets have not hit the centre of plate but been hit at positions very close to each
other repeatedly.
8. Hysteresis
According to the second law of thermodynamics, no process can be exactly reversible. The
instrument may give one output when the input is being increased slowly upto the peak
value, and another output for the same input, when it is decreased back to zero value. For
example, consider an LVDT that is used to measure small displacement (input) and gives
voltage as output. The input-output curve may be as shown in figure 2.3.
+ Voltage output

V2
V1

Maximum output
Hysteresis +
– Dead Band Displacement
(Input)

Maximum
Input hysteresis

Fig. 2.3 Hysteresis curve.

This error induced in the output for a given value of input when it is approached from
ascending order and descending order is called hysteresis. It is caused by magnetic
characteristics, backlash, friction, elastic deformation etc. Hysteresis error can be minimised
by taking output reading both for ascending and descending values of input and then taking
average, i.e.,
V1 + V 2
V average =
2
Dead band or dead zone is the case of hysteresis, which may be defined as total range
of input values for a given output. Numerically, it is twice of hysteresis.
9. Drift
Sometimes, the output of an instrument gets distorted due to certain interfering signal
 50 Principles and Practices of Measurement, Metrology and Control

(noise) such as temperature changes, component malfunctioning etc. This sensitivity of the
instrument to the interfering signals is called drift. It can be of two types :
(A) Zero drift
(B) Sensitivity drift.
Zero drift refers to shift in zero point because of noise signals, though the sensitivity
remains constant. Sensitivity drift refers to change in sensitivity of the instrument due to
noise signals. Refer figure 2.4.

O/P O/P O/P Measured Value

Measured Value Measured Value r

rro
tale
To
True Value True Value True Value
Zero Zero Zero
Drift Drift Sensitivity Drift Drift
I/P I/P I/P

Fig. 2.4 Drift in measuring instruments

MISCELLANEOUS PROBLEMS ON STATIC CHARACTERISTICS

EXAMPLE 2.1. For a spring balance, the force deflection data for the spring is tabulated
below :
Force (N) 0 10 20 30 40 50 60 70 80
Deflection (mm) 0 2 4 7 10 14 16 21 25

Find the best fit line to this data. Also find out linearity and sensitivity of the measuring
instrument.
Solution: Let the equation for best fit line be y = mx + C. Then according to the principle
of least square
y = mx + nC ...(A)
xy = mx2 + Cx ...(B)
x y x2 xy y’ y – y’

0 0 0 0 0 0
2 10 4 20 11.7 – 1.70
4 20 16 80 17.99 2.01
7 30 49 210 27.42 2.58
10 40 100 400 36.85 3.15
14 50 196 700 49.43 0.57
16 60 256 960 55.71 4.29
21 70 441 1470 71.43 – 1.43
25 80 625 2000 84.01 – 4.01
x = 99 y = 360 x2 = 1681 xy = 5840
 Static Performance Characteristics and Errors in Measuring Instruments 51
Substituting the value of x, y, x2,
xy from the table in equation (A) and (B), the
equations can be rewritten as
360 = m(99) + C × 9
5840 = m1681 + C × 99
This gives m = 3.1438 and C = 5.418. Hence the best fit line to the data is y = 3.1438x
+ 5.418
Sensitivity = Slope of line = 3.1438
M a x D evia t ion fr om b es t fit lin e
Percentage linearity = × 100
F u ll s ca le r ea d in g
4.29
= × 100 = 5.10%.
84.01
EXAMPLE 2.2. The equation of the best fit line for the following data is y = 0.672x + 0.296.
Comment upon the linearity of the measuring instrument.

x 0.9 2.2 3.3 4.5 5.7 6.7

y 1.1 1.6 2.6 3.2 4.0 5.0

Solution: For the values of x (independent variable), the corresponding value of y’ (dependant
variable) which lies on best fit line where y’ = 0.672x + 0.296 are as follows :

x 0.9 2.2 3.3 4.5 5.7 6.7

y’ 0.9 1.77 2.51 3.32 4.13 4.79

Let us draw a graph which shows the best fit line as well as the (x, y) coordinates.
y
5

1 2 3 4 5 6 7 8 x

Fig. 2.5.

From the graph, it can be seen that the observed readings are deviating from the least
square fit line. Very few points lie on the line. Hence, the instrument cannot be called
perfectly linear.

EXAMPLE 2.3. A measuring instrument consists of a transducer, an amplifier and a recorder.

Their individual sensitivities are 0.35 mV/ ° C, 3.0 V/mV and 6.6 mm/V respectively. Determine
the overall sensitivity of the instrument.
Solution : Let us make block diagram for this instrument.
 52 Principles and Practices of Measurement, Metrology and Control

Temperature Transducer Voltage Amplifier Voltage Recorder Displacement

(Input) 1 Kt = 0.35 mV/°C 2 Ka = 3.0 V/mV 3 Kr = 6.6 mm/V (output) 4

Fig. 2.6.

If Kt, K a and Kr be the sensitivities of transducer, amplifier and recorder, then

q2 q q
Kt = , Ka = 3 , K r = 4
q1 q2 q3
q4 q2 q3 q4
Overall sensitivity = q = q × q × q
1 1 2 1
= Kt × K a × K r
= 0.35 × 3 × 6.6 = 6.93 mm/ ° C.

EXAMPLE 2.4. A Wheatstone bridge requires a change of 10  in the unknown arm of the
bridge to produce a change of 4 mm in the galvanometer. Determine the sensitivity of the
Wheatstone bridge. Also determine the scale factor.
Ch a n ge in ou t pu t 4
Solution : Sensitivity = = = 0.4 mm/
Ch a n ge in in pu t 10
Ch a n ge in in pu t 10
Scale factor = = = 2.5 /mm.
Ch a n ge in ou t pu t 4

EXAMPLE 2.5. A mercury thermometer has a capillary tube of 0.35 mm diameter. What
should be the length of the tube in order to have sensitivity 0.35 mm per°C. Assume coefficient
of thermal expansion 1.81 × 10 –3 per°C.
Solution : We know that change in length per unit rise in temperature is given by l =
.L.t, where  is the coefficient of thermal expansion, L is the original length and t is the
rise in temperature.
Dl 0 . 35
Hence L = = mm
aD t 1 .81 ´ 10 –3
350
= mm = 194.4 mm = 19.44 cm.
1 .8
EXAMPLE 2.6. A moving coil voltmeter has a uniform scale with 100 divisions. The full
scale reading is 250 V and 1/10th of a scale division can be estimated with a fair degree of
certainty. Determine the resolution of the instrument in volt.
Solution :
250
1 scale division = = 2.5 V
100
1
Resolution = th of a scale division
10
1
= × 2.5 = 0.25 V
10
 Static Performance Characteristics and Errors in Measuring Instruments 53

2.2 ERROR
An error can be defined as difference between measured value and true value. An ideal
measuring instrument should be designed and manufactured for precise and unambiguous
measurement of parameters of interest. But since hundred percent control is unattainable,
error does occur. The major sources of error that may contaminate the results are : the
measurer and the instrument itself. Sometimes, the measurer may suffer from temporary
factors like fatigue, boredom, anxiety or other distractions. Such factors limit the ability of
the measurer to take the measurements accurately and fully. Hunger, impatience, or general
variation in mood can also effect the readings. Any condition that creates strain on the user
during the measurement session can have serious effect on data collection. A defective
instrument can also cause errors in measurement. Thus, an error may be systematic or
random in nature. Experimental errors can thus be classified as systematic errors and random
errors. Systematic errors are the ones associated with particular measurement instruments
that produce consistent errors. For example, an improperly calibrated scale might give weights
that are always larger than the true values. Similarly, a person who reads thermometer with
level of eyes either above or below (not a same level) the mercury meniscus position will
introduce systematic error in his/her data. Random errors result from unknown and
unpredictable variations in experimental situations. These are also referred as accidental
errors and are sometimes beyond the control of the observer. Random errors may be caused
by fluctuations in temperature, mechanical vibration of experimental setup or estimates of
measurement readings by the observer. Avoiding systematic errors depends on the skill of
observer to detect and prevent or correct them. One can minimize the effect of random errors
by taking many readings and finding their average so that random fluctuations becomes
statistically insignificant. A study of errors, its types and statistical analysis of data collected
becomes important so as to identify and eliminate the sources of errors.

2.2.1 Types of Error

Errors can be classified on the basis of their origin as follows :
Instrument error : There can be some errors that are inherent in the instrument itself
due to its poor design or manufacturing. Such errors limit the accuracy of the measuring
instrument even when it is new. Examples of instrument errors are mechanical friction,
wear, backlash, inertia effects, pointer drag, errors in divisions of graduated scale, in equality
of balanced arms, irregular spring tension, in appropriate tolerances on mating parts during
the design process, errors in assembling various parts of the instrument, electrical leakage,
poor optics etc. Instrument errors can be avoided or minimized by selecting suitable measuring
instrument and applying suitable error correction after determining the amount of instrument
error. Instrument should be calibrated against a suitable standard, periodically. Preventive
maintenance of the instrument should be taken up as per the schedule.
Environmental errors : Such errors are induced due to changing environmental conditions
of the measuring instrument. Environmental conditions that can significantly effect the
accuracy of an instrument are : changes in surrounding temperature, atmospheric pressure,
humidity, magnetic field etc. For example, if mercury-in glass thermometer is used to measure
room temperature but if the thermometer is placed near a window or in the sunshine, the
readings may be spurious. Changes in ambient temperature causes thermal expansion or
contraction of metallic measuring scale. A voltmeter may not give accurate reading if effected
by some external magnetic field. To eliminate environmental errors, an instrument should
be used in ambient conditions for which it is designed, manufactured and calibrated. This
may require controlling the ambient temperature, pressure or humidity etc. Alternately, a
 54 Principles and Practices of Measurement, Metrology and Control

new calibration may be done under the local conditions. Some times, it becomes necessary
to measure deviations in local conditions as compared to ideal environmental conditions so
as to apply suitable error correction.
Operational error : The errors induced in the readings if the instrument is not operated
properly are called operational errors. For example, a mercury-in-glass thermometer will
give accurate reading if the sensing bulb is not immersed in the working fluid completely.
A stroboscope may not give accurate readings of rpm if it is not operated in a dark room.
Vibrations, mechanical shock or electrical disturbance can also introduce errors.
Observational errors : The readings or observations of the measuring instrument must
be taken very carefully in order to avoid observational errors. Parallex is the most common
sort of observational error. It is the apparent displacement on the scale when the line of
vision is not normal to the scale. On a continuous scale, if the pointer measures in between
two consecutive (graduations) two marks, lack of ability to interpolate accurately leads to
error in the readings. Human errors such as carelessness or inexperience can lead to errors
being induced in the observations. Personal bias such as tendency to read the observed value
either high or low, tendency to take reading too early even before the initial oscillation die
out leads to systematic error being induced in the observations.
System interaction/loading errors : The process of testing and measurement can effect
the conditions of physical system under study. The errors so induced are commonly known
as loading errors. Some examples of loading errors are enumerated below.
(i) During temperature measurement, introduction of a thermometer into the system
leads to change in thermal capacity of the system and provides an extra path for heat
leakage, thus changing the original state of the system. This introduces loading error
in temperature measurement.
(ii) Reading shown by hand held tachometer can vary with variation in pressure with
which it is pressed against the shaft.
(iii) A millivoltmeter or ammeter will introduce additional resistance in an electrical
circuit thereby altering the circuit current by a significant amount.

2.2.2 Statistical Analysis

Sometimes, the measuring instrument may have a low precision, that is, on application
of same input repeatedly, the output of the instrument may not be exactly the same value
each time. The repeated readings tend to produce a band of results scattered about a point
or line of true values. Scatter plots can be used to understand the relationship between the
input and the output of the measuring instrument. They provide a means for visual inspection
of data which is easier to interpret as compared to tabulated list of input and output values.
But the most important objective of statistical analysis is to determine a single value for the
entire chunk of data so that it describes the overall level of the group of observations and
can be considered a representative value of the whole set of data. There are several measures
which give an idea of centre of distribution of data. We shall restrict ourself to only three
measures of central tendency namely arithmetic mean, median, mode.
Arithmetic mean : It is also referred as arithmetic average and is denoted by X . In
general, if there are ‘n’ values in the sample, then
x1 + x2 + x3 + .... + x n
X =
n
n
S xi
i =1
= where i = 1, 2, 3, ... n
n
 Static Performance Characteristics and Errors in Measuring Instruments 55
Note that if we subtract all the individual values from their mean, then some values will
be negative and some positive. If all these differences are added together, then the total sum
will be zero. Mathematically,
n
S ( xi - x ) = 0 for i = 1, 2, 3, ... n
i- 1

It can be contemplated that the mean is very sensitive to extreme values. Computation
of the arithmetic mean is based on all the values in the data. So, an extreme value in the
data would shift the mean towards it, thus making the mean value unrepresentative of the
data.
M e di an
It is a measure of central tendency. Median value appears in the centre of the ordered
data. Thus, it divides the list of ordered values in the data into two equal parts such that
half the data values will be less than the median and half values will be greater than the
median. If the total number of values in the data is odd, then after arranging the data in
increasing or decreasing order, the middle value gives the median. But if total number of
values in the data is even, the median is computed by taking the average of middle two
values.
M o de
It is defined as the most frequently occurring value in the data. It can be observed that
mode is not effected by extreme values in the data and can be easily obtained from an
ordered set of data. When every value has an equal number of observations, there is no
mode.

EXAMPLE 2.7. In a piston manufacturing company, five pistons were randomly picked up
and their outer diameter was found to be 50.19, 50.20, 50.24, 50.24 and 50.17 in mm. Find
the mean, median and mode.
SX
Solution: Arithmetic mean = X = = 50.208 mm
n
To find median, arrange the data in increasing order
50.17, 50.19, 50.20, 50.24, 50.24, hence median = 50.20
Mode = 50.24 mm
Mean, median or mode donot describe the data entirely. They provide point estimates of
the centre of distribution. For example, if the average depth of a lake is 5 feet it does not
give us confidence to walk through it. We must know how the data is distributed about the
mean. In other words, there should be a measure that desirable the spread of entire set of
measurement readings. In this text, we shall discuss three measures of dispersion namely
range, variance and standard deviation.
Range is defined statistically as absolute difference between the highest and the lowest
measured reading in the data set. In the example discussed above range is 50.24 – 50.17 =
0.07 mm.
Variance is one of the most useful measures of dispersion. It take into account each value
of the data and also as to how each observation is distributed. In general we can say that
if each value of collected measurement readings is reasonably close to the mean value, there
is very little dispersion of the data. On one other hand if the data is quite widely spread
(dispersed) and is at considerable distance from the arithmetic mean, then we can say that
 56 Principles and Practices of Measurement, Metrology and Control

data is highly variable. Variance is a measure of such variability. It is usually denoted by 2.
If N readings are taken for a measurement say X1, X 2, X3 ....XN, then Variance
N
S ( X i - X )2
2 = i =1
N
where X is the arithmetic mean and N is the number of observation which also is
referred as sample size.
Standard deviation is simply the square root of variance. It is denoted by sigma ().
Hence
N
S ( X i - X )2
= i =1
N
Probability Distribution
It is the listing of all possible outcomes of an experiment together with their probabilities.
It has been classified as discrete distribution and continuous distribution. Binomial distribution
is an example of discrete type where as normal distribution is an example of continuous
distribution. In this text, we shall restrict ourself to normal distribution. It is also important
due to the fact that in practice the experimental results very often seem to follow the normal
distribution. It is symmetrical, bell shaped curve as shown in figure. Normal distribution is
defined in terms of its mean () and standard deviation ()
y

Symmetrical
about mean

Tail extends
indefinitely

u x
Fig. 2.7. Normal distribution curve.

Height Y at any value of X is given by the relation

2
1 æx - mö
1 - ç ÷
2è s ø
Y = e
2 ps
An interesting fact to note is that the area under the curve between the mean and any
point x which is one standard deviation away from the mean is always the same no matter
what the values of  and  may be. Same is true for any point x which is two standard
deviation away and so on. This characteristic can be stated in general form as the number
of standard deviations Z for an observation, which is the distance between the value x and
the mean in terms of number of standard deviations in the normal distribution is defined by
x- m
Z =
s
 Static Performance Characteristics and Errors in Measuring Instruments 57
Z is known as the Z-score or standard normal deviation. The value of Z follows a normal
probability distribution with a mean of zero and standard deviation of one. This probability
distribution is known as standard normal probability distribution. A table of areas can be
used for all types of normal distributions.

n n n n n EXERCISE n n n n n

1. What is the significance of performance characteristics of measuring instruments?

2. How can you classify performance characteristics of measuring instruments.
3. Explain the following terms :
(i) Linearity
(ii) Static sensitivity
(iii) Span
(iv) Threshold
(v) Precision
4. Differentiate between :
(i) Threshold and resolution
(ii) Accuracy and precision
5. Draw the hysteresis curve.
6. What is drift?
7. Differentiate between zero drift and sensitivity drift.
8. Discuss the performance characteristics of LVDT.
9. Differentiate between range and span.
10. What are the potential sources of errors in measuring instruments.
11. Why is it essential to provide proper working environment for measuring instruments.
12. What are various types of errors in measuring instruments?
13. How are errors minimised using statistical analysis?
14. Explain the following terms :
(i) Zero drift and sensitivity drift
(ii) Precision and accuracy. (UPTU 2005-06)
15. Distinguish (any two of the following)
(i) Systematic and random errors
(ii) Passive and active transducers
(iii) X-Y and Y-t recorders
(iv) Accuracy and precision of an instrument. (UPTU 2006-07)

rrr