This article has been accepted for publication in a future issue of this journal, but has not been

fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/JSEN.2015.2442651, IEEE Sensors Journal

> Accepted for Publication in IEEE Sensors Journal (Ref. No.Sensors-12139-2015) < 1

Design and Implementation of Real-time Flow

Measurement System using Hall Probe Sensor
and PC based SCADA
S. Sinha, D.Banerjee, N. Mandal, Member, IEEE, R. Sarkar, Member, IEEE and S. C. Bera, Member,
IEEE

type, ultrasonic type, coriolis-force type, thermal or

Abstract—A rotameter is a variable area type flow rate anemometric type etc. In orifice type flow meter, an orifice
measuring instrument in which the position of a metallic float in plate combined with differential pressure transmitters (DPT)
a transparent conical tube is taken as a flow rate indicator. It has is used as an obstruction type volume flow rate transducer
the disadvantage that it is a local indicating type instrument and
special type of transducer is used for its remote indication. In the
used in both liquid and gas applications. In electromagnetic
present paper, a non-contact flow rate measurement technique flow meter, the induced voltage between two diametrically
using Hall probe sensor and rotameter is designed, developed opposite electrodes in contact with a conducting liquid
and tested. In this design, a float carrying a thin circular flowing through a pipeline section under an intense magnetic
permanent magnet is used and a Hall probe sensor placed outside field is directly proportional to the volume flow rate of the
the rotameter tube has been used to sense the variation of liquid. The turbine flow meter converts the velocity of flowing
magnetic field of the magnet with the variation of float position.
A signal conditioner unit has been used to convert the Hall probe fluid in a pipeline into analog or pulsating signal proportional
sensor output into 1-5 volt dc signal. This dc signal output of the to volume flow rate. The ultrasonic flow meter measures the
signal conditioner has been sent to a PC based flow indicator volume flow rate of a fluid from the Doppler effect of
through opto-isolator and analog input channel of a data ultrasonic wave in the upstream and downstream side of
acquisition system (DAS) card. The PC based flow indicator has flowing fluid between transmitter and receiver. There are two
been designed using Lab Tech Note Book Pro software and PC more very common ultrasonic flow meters based on vortex
based supervisory control and data acquisition (SCADA) system.
A theoretical equation has been derived to explain the operation principle and on changes in transit time caused by the velocity
of the system. The performance of the system has been tested of the intervening medium. Vortex flow meter can measure
experimentally and the experimental results are reported in the flow rate accurately under turbulent condition. Development
paper. A very good repeatability and linearity of results has been of smart and intelligent flow meters like coriolis mass flow
observed. meters, multi variable flow meters, microprocessor-based flow
meters [1-6] has also become popular in industry. The thermal
Graphical Abstract:
or anemometric type flow meter is also a mass flow meter, but
is of simpler design than coriolis flow meter. It works on the
principle of heat transfer by a flowing fluid from the hot
surface of a sensing element like hot wire resistance, diode
junction etc. F. Cascetta et al [8] have compared the
performances of two different types of insertion flow meter,
namely magnetic flow meter and turbine flow meter and have
shown that magnetic flow meter has the claimed accuracy in
the entire flow range but the other has the claimed accuracy
Index Terms—Flow Measurement, Hall Probe, Process only in upper range. F.Cascetta [9] has studied the effect of
Monitoring, PC based SCADA, Rotameter. pressure on the performance of a coriolis mass flow meter and
has shown that the fluid pressure affects the accuracy of
coriolis flow meter. S.G.Etemad [10] have shown that the
I. INTRODUCTION power-law index, the Reynolds number, and the distance

M EASUREMENT of flow rate through a pipeline is very

important in any process industry. The operation of a
flow measurement system primarily depends on its flow
between the impact and static tubes affect the operation of
pitot tube in flow measurement and have proposed correction
technique for both Newtonian and non-newtonian fluids by
transducer. The conventional flow measurement systems using neural network model. C.Clark et al [11] have designed
widely used in process industries may be of various types a modified straight tube coriolis flow meter with driving
depending on the types of transducers [1-5] such as frequency 750Hz which has reduced delay time and ultra fast
obstruction type, turbine type, electromagnetic type, vortex response. Thomas Povey et al [12] have demonstrated a new

1530-437X (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/JSEN.2015.2442651, IEEE Sensors Journal

> Accepted for Publication in IEEE Sensors Journal (Ref. No.Sensors-12139-2015) < 2

technique of mass flow rate measurement with reasonably due to initial mechanical stress has been reduced by using
high accuracy without the need for calibrated flow meters. In spinning current principle and have given a general discussion
this technique, the gas under measurement is allowed to exit [17] on development of the new Hall effect sensor. N. Mandal
through an initially pressurized plenum and its mass flow rate et al [19] have designed such a system using rotameter and
is determined from temperature and pressure measurement. R. Hall probe sensor.
Abdul Rahim et al [13] have developed an optical fiber based In the present paper, a modified technique of rotameter
multiple fan beam optical tomography to measure mass flow transducer has been designed and developed in order to avoid
rate of a fluid. U.R. Prasanna et al [15] have proposed a novel the difficulties of conventional LVDT type and inductance
method of fluid flow measurement by compensating the type rotameter transducer. In this technique, a thin circular
pressure drop across the ends of measuring unit using a permanent magnet is placed inside the float of the Rotameter
compensating pump. This technique is very accurate since null and a Hall probe sensor measures the variation of magnetic
deflection type measurement is used. R. R. Rhinehart et al field due to the movement of the float. The float along with
[16] have proposed a power law approach for online or offline the magnet changes its equilibrium position linearly with the
calibration of orifice type flow meters. S. C. Bera et al [18] change of flow rate with respect to the Hall probe sensor
have proposed a flow measurement and control technique mounted on the outside top surface of the rotameter tube.
using thyristor operated pump as final control element and Therefore, with the increase of flow rate, the float rises along
motor current as the flow sensing parameter. In this technique, the rotameter tube and the distance between the float magnet
it has been shown that the load current of the electric motor in and the Hall probe sensor decreases with the increase of
an electric motor operated pump is linearly related with flow magnetic field intensity sensed by the Hall sensor. Hence, the
rate of a liquid produced by the pump through a pipeline when Hall sensor output increases with the increase of flow rate. In
a tuned flow control loop is used. Qi-Li Hou et al [20] have order to design a PC based flow indicator the Hall sensor
described a digital drive and signal processing technique to output has been amplified by a signal conditioner circuit and
design and develop a coriolis mass flow meter. In this sent to PC through opto-isolator and DAS card. A virtual flow
technique, the stalling problem of coriolis flow meter tube has indicator has been designed using Labtech Note Book Pro
been eliminated by using positive-negative step signal for flow Software and PC based SCADA. A theoretical equation has
tube oscillation digital zero crossing detection technique has been derived to explain the performance of the developed flow
been used for measurement frequency and phase difference of indicating system. The proposed flow transducer and indicator
sensor output signals. MEMS type flow sensor [22] is a low system has been experimentally tested. Experimental results
cost electronic flow sensor with high accuracy and reliability are reported in the paper and have been found to follow the
and has gained popularity. P. Saccomandi [23] et al have theoretical equation with good linearity and repeatability.
described an opto electronic bidirectional air flow sensor in
infant artificial ventilation where two photodiodes combined II. METHOD OF APPROACH
with a T shaped target in differential modes are used as flow The rotameter is variable obstruction type flow measuring
sensor. E. Schena [24] et al have utilized the micro bending of instrument, which consists of a conical transparent tube placed
an optical fiber placed inside the flow tube combined with in vertical position in a pipeline through which fluid is
LED and quadrant position detector to design a flow sensor flowing. The tube consists of float, which is a solid inverted
for measuring low flow rates. conical object of density greater than the fluid, so that at no
Rotameter [1-5] is one type of variable area type flow meter flow it takes its minimum position closing the inlet pipe.
wherein the height of the float in a transparent vertical conical When fluid flows through the pipeline, the float takes an
tube is linearly related with the flow rate of fluid through the equilibrium position under the action of three forces [1-5],
tube. But it is generally used as a local indicator and has namely downward gravitational force, upward buoyant force
number of constraints such as vertical mounting, transparent and upward drag force. Under this condition, the height of the
tube and viscosity effect etc. It requires special type of upper surface of the float from its position at zero flow is
transducers for its remote indication. Linear variable linearly related with flow rate.
differential transformer (LVDT) is generally used in Let the float of rotameter is attached with a circular
transmitting rotameter reading to a remote location. However, permanent magnet as shown in Fig.1 (a) so that the magnet
LVDT type rotameter transducer has very small range in moves upwards with the float when flow rate of a fluid
measuring the displacement of rotameter float in flow increases. Therefore, magnetic field intensity at an axial point
measurement. This type of rotameter transducer may have also of the ring magnet increases with increase of flow rate. Let a
many other disadvantages such as requirement of ac excitation Hall sensor is placed at this axial point outside the rotameter
at high frequency, requirement of phase sensitive rectifier, tube to measure this variation of magnetic field intensity as
error due to leakage flux, hysteresis loss and eddy current loss. shown in Fig.1 (a) and Fig.2. Now to understand the variation
N. Mandal et al [21] have designed and tested a flow of magnetic field at an axial point of a ring magnet, an
transmitter using a direct self-inductance type Rotameter experimental set-up as shown in Fig.1 (b) is made.
transducer. Hall probe sensor may be used to measure the flow
rate of the rotameter. V. N. Petoussis et al [14] have
developed a new Hall effect sensor where the effect of offset

1530-437X (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/JSEN.2015.2442651, IEEE Sensors Journal

> Accepted for Publication

ublication in IEEE Sensors Journal (Ref. No.Sensors-12139
12139-2015) < 3

and the corresponding value of axial magnetic field is

measured by the digital gauss meter. The static characteristic
of ring magnet obtained by plotting magnetic field (B) against
agains
axial distance (x)) is shown in Fig.1(c).
600

500

400
Best fit
Measurd Value

B [G]
300

200

Y=540.1X -1.99
100

0
0 2 4 6 8 10 12 14 16 18
Axial Distance (x) [cm]
Fig. 1(c). Static characteristic of ring magnet from experiment.
experiment

From Fig. 1(c), it is observed that the axial magnetic field (B)
of a ring magnet almost follows an inverse square root relation
above certain limiting
imiting value of axial distance(x). Hence, this
Fig. 1(a).Diagram
Diagram of float with ring magnet and Hall probe sensor
sensor. may be approximately given by empirical equation,

K1
Bx ≈ (1)
x2

As the flow rate increases, the float along with the magnet will
move upwards in the rotameter tube and the magnet will come
closer to the Hall probe placed on the outside top surface of
the rotameter tube as shown in Fig. 2.

Fig. 1(b).Experimental set-up

up for measurement of axial magnetic field of a
ring magnet.

In Fig.1 (b) a Hall probe-based

based three digit digital gauss Fig. 2.Diagram
Diagram of the proposed transducer along with float and Hall probe
sensor .
meter (Make; M/s Scientific Equipment & Services (ISO
9001:2000 certified), Model: DGM-169,169, Sl. No.: 89, Roorke,
Let the flow rate through the pipeline be Q and the
India), has been used to measure the magnetic field produced
corresponding
sponding height of the float be h. We know from the
by a ring magnet of inner radius ‘a’.’. The Hall sensor of the
principle of rotameter that flow rate is directly proportional to
gauss meter is placed at an axial pointt ‘P’
‘ with its surface
float height. Hence,
perpendicular to the axis ‘OP’ of ring magnet, along which its
magnetic field acts. The ring magnet used in the experiment is
Q = K2h (2)
of outer diameter 53 mm, inner diameter 24 mm and thickness
10 mm. The axial distance ‘OP’ of this Hall sensor
sen from the
centre ‘O’ of top surface of ring magnet is measured by a scale where K 2 is a constant.

1530-437X (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/JSEN.2015.2442651, IEEE Sensors Journal

> Accepted for Publication

ublication in IEEE Sensors Journal (Ref. No.Sensors-12139
12139-2015) < 4

Now the distance of the Hall probe from the float magnet is K (8)
Q = K2H − 1
2
x = H −h (3) Vh
1
where, K = K 2 ( K 4 ) 2 (9)
where H is the distance of the Hall sensor from the float
surface at its minimum position at no flow. Hence, from Assuming, y = 1 , the equation (8) is reduced to
1
equations (1) and (3) we have,
Vh 2
K1 K1 Q = K2H − K y (10)
Bx = = (4)
x 2 ( H − h )2 or, y = − aQ + b (11)
Now the Hall voltage Vh of the Hall
all sensor is proportional to
where, a = 1 & b = K 2 H (12)
Bx if the current passing through the sensor be kept constant. K K
Hence, we have the following equation. The signal conditioner circuit using INA101 has very low
noise instrumentation amplifier is shown in Fig. 3.
Vh = K 3 B x (5)

where K3 is the constant of proportionality.

From equations (2), (4) and (5) we have,

K4 (6)
Vh = 2
 Q 
H − 
 K2 

where K 4 = K1 K3 = constant.
Fig. 3.Circuit
Circuit diagram of Hall sensor based flow transducer.
transducer

Thus, the output (Vh ) of Hall sensor is nonlinearly related Let the signal conditioner circuit be calibrated for flow rate
with volume flow rate (Q). in the range from Q1 L/min (liter per minute) to Q2 L/min for
Since at low flow rate, the float height (h) is small so we may
which the sensor output varies from Vh1 mV to Vh 2 mV and
assume that H ≫ h . Hence, the equation (6) may be
the signal conditioner output changes from 1 volt to 5 volt.
approximated as
Hence, the gain of the signal conditioner circuit is given by,

K4  2Q  (7)
Vh = 1 +  K′ =
4 (13)
H2  K 2H  V h 2 − V h1

Thus at low flow, the Hall sensor output

put is linearly related If V0 be the signal conditioner output in volts for flow rate Q
with flow rate. L/min and Hall sensor output Vh mV then Vh is given by,

III. DESIGN V0 − 1
Vh = + V h1 (14)
In the present design, a circular permanent mag magnet is K′
selected of inner radius a = 11mm , depth d = 8mm , and width
w = 11mm as shown in Fig.1 and value of H in Fig.2 is From equations (11) & (14), we have,
250mm. The Hall sensor used in this work is a Solid-state
Solid
effect sensor (SS490) with size 0.160 × 0.118 inch,
Hall-effect V0 − 1
b − [1 / ( + V h1 )]1/ 2
having power consumption of 7mA at 5V DC. It has a positive
posit Q= K′ (15)
temperature coefficient of (+0.02% / 0 C ) with accuracy < 3% a

and operating temperature −400 C to +1500 C . The float is made

of stainless steel. IV. EXPERIMENT
Experiment is performed in two steps with the experimental
From equation (6) we have, set up as shown in Fig. 4. Here, tap water is used as a process
fluid at room temperature. The set up consists of an
underground tank and an overhead tank each of 300-litre300

1530-437X (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/JSEN.2015.2442651, IEEE Sensors Journal

> Accepted for Publication in IEEE Sensors Journal (Ref. No.Sensors-12139-2015) < 5

capacity. The tanks are connected with 25 mm diameter TABLE I

EXPERIMENTAL DATA FOR STATIC CHARACTERISTIC OF FLOW TRANSDUCER
pipeline through 25 mm ball valve, rotameter (2-24 L/min)
and a pump (1φ, 230 V, 370 watt).
Sl Flow Hall Probe voltage (mV)
No rate Set 1 Set 2 Set 3
. (L/min) Increa Decre Increa Decreasi Increas Decrea
sing asing sing ng ing sing
1 2 1.51 1.5 1.5 1.51 1.51 1.5
2 4 2.01 2.01 2.01 2.01 2.01 2.01
3 6 2.51 2.51 2.5 2.5 2.51 2.5
4 8 3.11 3.11 3.11 3.11 3.11 3.11
5 10 4.22 4.22 4.22 4.22 4.21 4.22
6 12 6.07 6.07 6.07 6.07 6.07 6.07
7 13 7.47 7.48 7.48 7.48 7.48 7.48
8 14 10.75 10.75 10.75 10.75 10.75 10.75
9 16 16.73 16.73 16.73 16.73 16.73 16.73
10 18 37.28 37.28 37.28 37.28 37.28 37.28
11 20 61.01 61 61 61.01 61 61
12 23 144.9 144.9 144.9 144.9 144.9 144.9
13 24 200 200 200.1 200.1 200 200

250

Increasing 1
Fig. 4.Experimental set up. 200 Decreasing 1
Increasing 2
Hall Sensor Voltage [mV]

Decreasing 2
In the first step, the characteristics of the proposed Hall 150 Increasing 3
sensor based rotameter transducer unit are determined and in Decreasing 3
the second step, the transducer is connected with PC through
100
signal conditioner, opto isolator and DAS card and the
corresponding static characteristic of the whole flow
indicating system is determined. To determine the 50

characteristic of the flow transducer, the flow rate through the

pipeline is increased in steps by opening the inlet valve (V1) 0
0 5 10 15 20 25
and at each step the Hall sensor output is measured by digital Q [L/min]

multimeter. The characteristic graph obtained by plotting Hall

Fig. 5.Characteristic graph obtained by plotting Hall voltage against flow rate.
sensor output against flow rate is shown in Fig. 5. Experiment
was repeated both in increasing and decreasing modes for six -3
x 10
times and the standard deviation curve for six observations is 6
shown in Fig.6.The experimental data are shown in Table-I.
The linearization characteristics obtained by plotting inverse 5
square root of Hall sensor output voltage against flow rates are
shown in Fig.7 (a) and 7(c). The percentage deviation from 4
S tandard Deviation

linearity of the characteristic of Fig. 7(a) is shown in Fig.7 (b).

In Fig. 7(a), the inverse square root characteristic for average 3
value of Hall sensor outputs in three increasing and three
decreasing modes is shown and in Fig.7(c) the characteristics
2
for three increasing and three decreasing modes are shown.
From the experimental graph, shown in Fig.7(a), it is observed
1
that the reciprocal of square root of Hall sensor output voltage
is related with flow rate by the following equation.
0
0 5 10 15 20 25

y = − 0.04 Q + 0.8807
Q [L/min]
(16)
Fig. 6.Standard deviation curve of the Hall Probe based flow transducer.
Comparing (11) and (16) we have a = 0.04 V −0.5 min/ L and
b = 0.8807 V −0.5 .

Hence, Q = 0.8807 − (1 / V h )
1/ 2
(17)
0.04

1530-437X (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/JSEN.2015.2442651, IEEE Sensors Journal

> Accepted for Publication in IEEE Sensors Journal (Ref. No.Sensors-12139-2015) < 6

0.9
Equation (18) is used in designing the PC based flow indicator
by using LAB Tech Note Book Pro software [14] as PC based
0.8
Measured
Best fit
SCADA. The virtual flow indicator is designed by using a
0.7
number of icon blocks selected from icon tray as shown by the
1/Sqrt Vh [1/Sqrt mV]

0.6 block diagram in Fig.8.

0.5

0.4

0.3

0.2

0.1
2 4 6 8 10 12 14 16 18
Q[L/min]

Fig. 7.(a).Linearization characteristic obtained by plotting inverse square root

of average Hall sensor output against flow rat.

1.5
Fig. 8.Block diagram of PC based signal-processing unit.

1
The flow rate in L/min along with input signal in volt is
% Deviation from linearity

0.5 displayed in vision icon block. Therefore, in the second step of

0 the experiment, the signal conditioner output in the range 1-5
-0.5
volt was connected to the designed PC based flow indicator
through opto isolator and DAS card. The static characteristic
-1
graph of the PC based indicator was then determined by
-1.5
increasing flow rate in steps and observing the indicated flow
-2
2 4 6 8 10 12 14 16 18
rate in PC monitor. This is as shown in Fig. 9(a) and the
Q [L/min]
experimental data are shown in Table-II. Corresponding
Fig. 7. (b).Percentage deviation from linearity of inverse square root percentage deviation curves from linearity, standard deviation
characteristic of Hall sensor.
curves from eight repeatable data and uncertainty curves are
shown in Fig. 9(b), Fig.9(c) and Fig. 9(d) respectively.
0.9

18
0.8

16
0.7 Increasing
Sensor Output in mV - 0.5
Inverse Square Root of Hall

Decreasing
14
0.6 Increasing
Decreasing
Increasing 12
0.5 Increasing1
Qo[L/min]

Decreasing Decreasing2
10 Increasing3
0.4
Decreasing4
8 Increasing5
0.3 Decreasing6
Increasing7
6
0.2 Decreasing8

4
0.1
2 4 6 8 10 12 14 16 18
Q [L/min] 2
2 4 6 8 10 12 14 16 18
Fig. 7.(c).Inverse square root characteristic of Hall sensor in increasing and Qa[L/min]

decreasing modes. Fig. 9.(a).Characteristic graph obtained by plotting the observed flow rate in
PC monitor against the true flow rate.
Now from experiment, it is observed that the average value 1.2

non-linear output (Vh ) of Hall sensor is in the range from Vh1 1

= 1.505 mV to Vh 2 = 37.28 mV when the flow rate changes 0.8

% Deviation from linearity

from 2 L/min to 18 L/min. Therefore, the gain of the signal 0.6

conditioner is obtained from designed equation (13) as 0.4

K′ = 111.81 0.2

For signal conditioner output voltage V0 , the corresponding 0

flow rate Q can be obtained from designed equation (17) as -0.2

-0.4

V0 − 1 -0.6
0.8807 − [1 / ( + 1.505 × 10 − 3 )]1/ 2 2 4 6 8 10
Q[L/min]
12 14 16 18

Q= 111.81 (18) Fig. 9.(b).Percentage deviation curves from linearity of the whole measuring
0.04 system.

1530-437X (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/JSEN.2015.2442651, IEEE Sensors Journal

> Accepted for Publication in IEEE Sensors Journal (Ref. No.Sensors-12139-2015) < 7

0.055
Fig. 9(a) has a good linearity. The percentage deviation curve
0.05
from linearity is shown in Fig.9 (b), standard deviation curve
0.045
for eight repetitive data is shown in Fig. 9(c) and standard
0.04 uncertainty curve is shown in Fig.9 (d). According to ASTM
International standard, repeatability limit is calculated by the
Standard Deviation

0.035

0.03 formula r = 2 2 Sr where Sr is the relative standard

0.025
deviation. It should lie within 5% of maximum range. From
0.02
the overall characteristic of the PC based measuring unit
0.015
shown in Fig. 9, the maximum percentage deviation curve
0.01
from linearity of the proposed instrument and standard
0.005
2 4 6 8 10
Q[L/min]
12 14 16 18 deviation curve are shown in Fig. 9(b) and Fig.9(c)
Fig. 9.(c)Standard deviation curves of the whole measuring system.
respectively. From these figures maximum percentage
deviation from linearity is found to be 1.12%. Also maximum
0.02
relative standard deviation is 0.053/17.95 = 0.00295265 =
0.018
0.295% with repeatability 0.834% and minimum relative
0.016
standard deviation is 0.0051/12.003 = 0.00042489 = 0.042%
0.014
with repeatability 0.118% where 17.95mV and 12.003mV are
0.012
Uncertainity

the respective average values after eight repetitions in

0.01

0.008
increasing and decreasing modes. Hence repeatability of the
0.006
proposed measuring unit lies within acceptable limits. Thus
0.004
the measuring unit may be considered to have industrially
0.002
acceptable repeatability. We know estimated standard
0
uncertainty (u) of measurement is calculated from the formula,
2 4 6 8 10 12 14 16 18
Q[L/min]
u=S n , where S is the standard deviation of an
Fig. 9.(d).Standard uncertainty curves of the whole measuring system.
experiments repeated for “n” times. From the overall
TABLE II characteristic of the proposed PC based flow indicator as
EXPERIMENTAL DATA FOR STATIC CHARACTERISTIC OF PC BASED FLOW shown in Fig.9(a), and the standard deviation curve shown in
INDICATOR Fig.9(c), a standard uncertainty curves has been drawn as
Sl Actu Observed (Q0) flow rate (L/min)
shown in Fig.9(d). From this curve, it may be concluded that
Sl. al
Set 1 Set 2 Set 3 Set 4
the proposed measuring system has a very small uncertainty (<
No. (Qa)
flow
2%). Similarly, from standard deviation curve of Hall sensor
Decr Decr Decr
rate Increas
easin
Incre
easin
Incre
easin
Incre Decre shown in Fig. 6, the repeatability and uncertainty of the Hall
(L/ ing asing asing asing asing
min)
g g g sensor may be calculated by following the similar procedure
1 2 2 2.01 2.01 2 2 2.02 2.02 2 and it is observed that the proposed transducer has a good
2 4 3.98 3.97 3.96 3.97 3.98 3.97 3.97 3.98
repeatability with small uncertainty.
The rotameter used in the present study is a commercial unit
3 6 5.82 5.81 5.82 5.81 5.81 5.82 5.81 5.82
with float made of stainless steel in an inverted glass tube with
4 8 8 8.01 8 8 8 8.02 8.01 8
specifications, ± 2% accuracy, ± 1% repeatability and range 2-
5 10 10 10 10.1 10.01 10 10 10.01 10 24 L/min. So all the data reported in the present study are
6 11 11.01 11.02 11 11.01 11 11.02 11 11 based on actual flow rate measured by rotameter with this
7 12 12 12 12.01 12 12 12.01 12 12.01 accuracy and repeatability. However, during experiment the
8 13 13 13.02 13.01 13 13 13.02 13.02 13 flow rate was obtained from a constant height overhead tank
9 14 13.7 13.72 13.72 13.7 13.7 13.72 13.72 13.7 for each reading of rotameter so that float position remains
10 16 15.78 15.77 15.77 15.76 15.78 15.79 15.77 15.78 stable at each reading. Again, before starting the experiment
11 18 18 18 17.9 17.9 17.9 17.9 18 18 the rotameter itself has been calibrated by using direct water
collection method. In this method the water passing through
rotameter for a stable float position is collected in a bucket for
V. DISCUSSION a certain interval of time, a measuring cylinder of laboratory
grade measures the volume, and then the volume flow rate is
The characteristic graph of the flow transducer as shown in
calculated from the ratio of collected volume to observed time.
Fig. 5 is nonlinear in nature and appears to follow the
This method has been repeated for each reading for three
characteristic equation (6). The characteristic graph obtained
increasing and three decreasing modes. The percentage error
by plotting inverse square root of Hall sensor output (1/ Vh ) between observed reading and actual reading obtained from
against flow rate (Q) is found to be quite linear as shown in calculation has been determined along with standard deviation
Fig. 7 and appears to follow the equation (11).The final static of six data. The percentage error has found to be very small
calibration graph of the PC based flow indicator as shown in less than ± 0.2% with very good repeatability.

1530-437X (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/JSEN.2015.2442651, IEEE Sensors Journal

> Accepted for Publication in IEEE Sensors Journal (Ref. No.Sensors-12139-2015) < 8

The conventional magnet type rotameter transducer output bears a good linear relation with flow rate with
generally consists of a magnetic float, which is magnetically percentage deviation from linearity lying almost within ± 2%
coupled with the shaft of a potentiometer. Thus, the float as shown in Fig.7 (b).
movement is converted into potentiometer movement and a PC based SCADA software used in the proposed
suitable mechanical arrangement is needed to transmit the monitoring system is Lab Tech Note Book Pro Software. The
rotameter reading in the form of current signal. This system data acquisition card used in this software is ADVANTECH
may suffer from error due to miscoupling between magnetic make, PCL-818L, 40 kS/s, 12bit, 16 channel, ISA
float and potentiometer head. The proposed transducer does multifunction card which can take both analog and digital
not suffer from this error with much more simpler design. This input with 0.1% accuracy for ± 10 volt and ± 5 volt analog
design is also very simple compared to that of LVDT type input ranges and 0.2% accuracy for ± 2.5 volt and ± 1.25 volt
rotameter. Hence, the cost of the transducer in commercial analog input ranges. The icon-based blocks available in this
form may be reduced to a small value compared to other software are used for signal processing of the analog input
conventional rotameter transducers. The novelty of the signal obtained from the output of the signal conditioning
proposed Hall sensor type rotameter transducer may be circuit of the proposed Hall probe sensor as shown in Fig.8. It
assumed to lie in the fact that the proposed transducer is a may be mentioned here that the same software or similar other
non-contact type rotameter transducer with better reliability software may also be used for designing a control loop.
and longevity compared to these conventional rotameter
transducers. The only disadvantage of the proposed sensor is VI. CONCLUSION
the non-linear output, which can be easily linearized by using
The design of the measurement system is very simple. The
any linearization technique that can be easily designed through Hall sensor & permanent magnet are now available at a very
any electronic circuit or microprocessor, or microcontroller or low cost, so the cost of the flow transducer will be low. The
PC based software technique. In the present paper, a PC based lifetime of the transducer will be high since the Hall sensor is
linearization technique has been described by using Lab Tech not in contact with the liquid and permanent magnet has long
Note Book Pro software. life stability. The PC based linearization technique of the
The operation of the proposed measuring system depends measuring circuit has been found to operate with good
on effect of temperature on Hall probe. Since the Hall probe is repeatability. The measuring system can be incorporated in a
kept in atmosphere and its temperature coefficient is very control loop with virtual controller designed by using the same
small so the effect of temperature on the net Hall voltage may software. Hence, it may be concluded that the proposed PC
be considered negligible for small change of atmospheric based flow indicator may be assumed to have good
temperature. However, for large change of temperature acceptability in industry in modern PC based Instrumentation
between winter and summer the sensor may be required to be system within certain limiting range of flow rate for which
recalibrated. To avoid this, bridge circuit based or IC based equation (1) is valid. The proposed measuring instrument may
temperature compensating circuit may be used or the sensor have some drawbacks such as limited range depending on size
may be kept in a constant temperature enclosure. of float and magnet, measurement error due to deviation from
It may be mentioned that only the linear portions of the inverse square law, shifting of position of Hall sensor from
graphs for flow rate upto 18 L/min have been shown in Fig.7 face to face position with float. The range of the measuring
(a) and Fig.7(c). The nonlinearity at higher flow rate may be instrument may be extended by modification of design of
rotameter.
due to the fact that at higher flow rate the distance between
float magnet and Hall sensor decreases to a small value for
ACKNOWLEDGMENT
which equation (1) is not valid. At low flow rate, this distance
is large for which equation (1) is valid. So during plotting of The authors are thankful to the All India Council of
the graphs shown in Figs. 7(a) and (b) obtained from Fig.5 the Technical Education (AICTE), MHRD, Govt. of India for
experimental data above 18 L/min flow rates are omitted. It is their financial assistance in the present investigation and the
observed that the linear graphs shown in Figs.7 (a) and (c), Department of Electronics Engineering, Indian School of
slightly deviates from linearity at some points in increasing Mines, Dhanbad, (Officially declared as IIT Dhanbad) for
and decreasing modes that may be due to human errors such as providing the facilities to carry out this work.
error in taking reading by digital mill voltmeter, error in taking
reading of rotameter, error in placement of Hall sensor etc.
There may be some other errors due to earth’s magnetic field REFERENCES
interference, electromagnetic interference etc. The graph [1] Bela, G. Liptak, “Process Measurement and Analysis”, 3rd ed.,
Butterworth Heinman Ltd., 1999.
shown in Fig.7(a) represents the average values of measured
[2] J. P. Bently, “Principles of Measurement systems”, 3rd ed., Longman
data in three increasing and three decreasing modes shown in Singapore Publishers Ltd., 1995.
Table-I and in Fig.7(c) the characteristic for three increasing [3] D. M. Considine, “Process Instruments and Control Hand Book”, 2nd
and decreasing modes are shown. It is observed that increasing ed., McGraw Hill Book company, 1974.
[4] R. C. Baker, “Flow Measurement Handbook: Industrial Designs,
and decreasing characteristics almost coincide with one Operating Principles, Performance and Applications”, Cambridge, U.K.,
another. All these graphs appear to show that within certain Cambridge Univ. Press, 2000.
limiting value of flow rate, inverse square root of Hall sensor

1530-437X (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/JSEN.2015.2442651, IEEE Sensors Journal

> Accepted for Publication in IEEE Sensors Journal (Ref. No.Sensors-12139-2015) < 9

[5] E. O. Doeblin, “Measurement system application and design”, 4th Sunita Sinha was born in India in 1986. She received the
edition, McGraw Hill Publishing Company, 1990. B.Tech degree in electronics and communication engineering
[6] US Department of the Interior Bureau of Reclamation, “Water
Measurement Manual” 3rd edition, Chapter-3. from MJP Rohilkhand University, Bareilly, India, in 2011 and
[7] “LabTech Data Acquisition and Process Control Software” (4/94 the M.Tech degree in electronics design and technology from
Release), Laboratory Technology Inc, 1994 edition. Tezpur Central University, Tezpur (Assam), India, in
[8] F. Cascetta “Effect of fluid pressure on coriolis mass flow meters 2013.She is currently pursuing the Ph.D(Engg.) degree with
performance” ISA Transactions, Elsevier, vol 35, Issue 4, 1996, pp. 365-
370. the Department of Electronics Engineering, Indian School of
[9] F. Cascetta, A. Palombo, G. Scalabrini, “Water flow measurement in Mines (Government of India, MHRD), Deemed University,
large bore pipes: An experimental comparison between two different Dhanbad, India. Her research interests include Transducer
types of insertion flow meters” ISA Transactions, Elsevier, vol. 42, Issue Development and controller design for process plant.
2, April 2003, pp 171–179.
[10] S.G. Etemad, J. Thibault, S.H. Hashemabadi “Calculation of the Pitot
tube correction factor for Newtonian and non-Newtonian fluids” ISA
Transactions, Elsevier, vol. 42, Issue 4, October 2003, pp. 505-512.
[11] C. Clark., M. Zamora., R. Cheesewright., and M. Henry., “The dynamic
performance of a new ultra-fast response Coriolis flow meter”, Flow
measurement and instrumentation, Elsevier, U.K, vol. 17, Issue 6, Dec.,
2006, pp. 315-414.
[12] Thomas Povey, Paul F. Beard, “A novel experimental technique for
accurate mass flow rate measurement” Flow Measurement and
Instrumentation, Elsevier, vol. 19, Issue 5, Oct. 2008, pp. 251-259. Deblina Banerjee was born in India in 1988.
[13] R. Abdul Rahim, L.C. Leong, K.S. Chan, M.H. Rahiman, J.F. Pang, She received the B.Tech. degree in Instrumentation and
“Real time mass flow rate measurement using multiple fan beam optical
tomography” ISA Transactions, Elsevier, vol. 47, Issue 1, January 2008,
Control Engineering from Haldia Institute of Technology
pp. 3–14. under WBUT, India, in 2009, and the M.Tech. degree in
[14] Vlassis N. Petoussis, Panos D. Dimitropoulos and George Stamoulis “A Control and Automation Engineering from Asansol
Novel Hall Effect Sensor Using Elaborate Offset Cancellation Method”, Engineering College, under WBUT India, in 2013. She is
Sensors & Transducers Journal, vol. 100, Issue 1, January 2009, pp. 85-
91.
currently pursuing the Ph.D. (Engg.) degree with the
[15] U.R. Prasanna, L. Umanand , “Non-disruptive and null-deflection mass Department of Electronics Engineering, Indian School of
flow measurement by a pressure compensation technique” Flow Mines (Government of India, MHRD), Deemed University,
Measurement and Instrumentation, Elsevier, vol. 21, Issue 1, March Dhanbad, India. Her current research interests include
2010, pp. 54-61.
[16] R. R. Rhinehart, S. Gebreyohannes, U. M. Sridhar, A. Patrachari, Md. S.
transducer development, calibration of transducer and
Rahaman, “A power law approach to orifice flow rate calibration” ISA controller design.
Transactions, Elsevier, vol. 50, Issue 2, April 2011, pp. 329–341.
[17] Vlassis N. Petoussis, Panos D. Dimitropoulos, and George Stamoulis,
“General Development of a New Hall Effect Sensor” Sensors &
Transducers Journal, vol. 127, Issue 4, April 2011, pp. 36-44.
[18] S.C. Bera, N. Mandal, and R. Sarkar, “A novel technique of using a
thyristor driven pump as the final control element and flow indicator of a
flow control loop” ISA Transactions, Elsevier, vol. 50, Issue 3, July
2011, pp. 496–503.
[19] K. Chakraborty, N. Mandal, and R. Sarkar, “Design of an Electronic
Flow Transmitter Using LVDT & Hall Sensor”, Int. Journal of
Electronics & Comm. Tech., vol. 4, Issue - 1, 2013. Nirupama Mandal (M’15) was born in
[20] Qi-Li Hou, Ke-Jun Xu, Min Fang, Cui Liu, Wen-Jun Xiong,
“Development of Coriolis mass flow meter with digital drive and signal India in 1978. She received the B.Sc. (Hons.) degree in
processing technology” ISA Transactions, Elsevier, vol. 52, Issue 5, physics, the B.Tech. degree in instrumentation engineering,
Sept. 2013, pp. 692–700. the M.Tech. degree in instrumentation and control
[21] N. Mandal, B. Kumar, R. Sarkar, and S. C. Bera, “Design of Flow engineering, and the Ph.D. (Tech.) degree in instrumentation
Transmitter Using an Improved Inductance Bridge Network and
Rotameter as a Sensor,” IEEE Trans. on Instrum. and Meas., vol. 63, engineering from the University of Calcutta, Kolkata, India, in
no.12, pp. 3127-3136, Dec. 2014. 2001, 2004, 2006, and 2012, respectively. She was an
[22] N. T. Nguyen, “Micromachined flow sensors - a review”. Flow Assistant Professor and the Head of the Department of
measurement and Instrumentation, Elsevier, vol. 8, Issue 1, May 1997, Electronics and Instrumentation Engineering at Asansol
pp. 7–16.
[23] P.Saccomandi, E. Schena, and S. Silvestri, “A novel target-type low Engineering College, Asansol, India, for eight and a half
pressure drop bidirectional optoelectronic air flow sensor for infant years, and joined the Department of Electronics Engineering at
artificial ventilation: Measurement principle and static calibration”, AIP, the Indian School of Mines (Government of India, MHRD),
Rev. of Sci. Instrum., vol. 82, no.2 (024301), Feb. 2011. Deemed University, Dhanbad, India, as an Assistant Professor
[24] E. Schena, P. Saccomandi, and S. Silvestri, “A high sensitivity fiber
optic macro-bend based gas flow rate transducer for low flow rates: in 2013. She has authored 15 papers in international journals
Theory, working principle, and static calibration” AIP, Rev. of Sci. and 15 papers in international and national conference
Instrum., vol. 84, no.2 (024301), Feb. 2013. proceedings. Her current research interests include transducer
design, PC-based instrumentation, controller design, and
process modeling.

1530-437X (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI
10.1109/JSEN.2015.2442651, IEEE Sensors Journal

> Accepted for Publication

ublication in IEEE Sensors Journal (Ref. No.Sensors-12139
12139-2015) < 10

Rajan Sarkar (M,14) was born in India in

1977. He received the B.E. degree in electrical engineering
from the National Institute of Technology, Agartala, India, in
1999, and the M.Tech. degree in instrumentation engineering
and the Ph.D. (Tech.) degree from the Univer
University of Calcutta,
Kolkata, India, in 2001 and 2012, respectively. He was a
Lecturer with the Bankura Unnayani Institute of Engineering,
Burdwan, India, from 2001 to 2002. He joined Asansol
Engineering College, Asansol, India, as a Lecturer in 2002,
where hee is currently an Associate Professor and the Head of
the Department of Electrical Engineering. Presently, two
scholars have registered under him for the Ph.D. (Engg.)
degree under the Indian School of Mines (Government of
India, MHRD), Deemed University, Dhanbad,anbad, India. He has
authored 14 papers in international journals and 10 papers in
international and national conference proceedings. His current
research interests include process transducers and process
process-
plant instrumentation.

Satish Chandra Bera(M,12)

Bera was born in
India in 1952. He received the B.Sc. (Hons.), B.Tech.,
M.Tech., and Ph.D. (Tech.) degrees in instrumentation
engineering from the University of Calcutta, Kolkata, India,
in1973, 1976, 1978, and 1988, respectively. He was an
Instrumentation Engineer with Hindustan Fertilizer
Corporation Ltd., Haldia, India, from 1982 to 1991, and joined
the University of Calcutta as a Faculty Member in 1991,
where he is currently a Professor of Instrumentation
Engineering with the Department off Applied Physics. He is
involved in carrying out a number of research projects on
transducer development, process-plant plant instrumentation,
process modeling, and biomedical instrumentation. He has
guided nine research scholars who have been awarded with the
Ph.D. (Tech.) degree from the University of Calcutta.
Presently, five scholars are working under him for the Ph.D.
(Tech.) degree. He holds four patents on transducer
development, and has applied for one more patent on steam
generation plant operation in a thermal power plant. He has
more than 140 papers in international and national journals
and conference proceedings.

1530-437X (c) 2015 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See
http://www.ieee.org/publications_standards/publications/rights/index.html for more information.