Experiment # 1 | Subject: Calibration of Displacement Transducer

Section #1 | Group 14

Mert Dilik 2016405186

Batuhan Burak Aydoğan 2016405072
Ahmet Taha Yayman 2014403024
Cüneyt Mert Cindoruk 2016405093
Emrullah Onur Dabakoğlu 2017405132

April 2021
TABLE OF CONTENTS

Abstract…………………………………………………………………………………………...2

Introduction………………………………………………………………………………………3

Theory……..………………………………………………………………………………………3

Setup………………………………………………………………………………………………5

Results…….………………………………………………………………………………………6

Analysis of Results………………………………………………………………………………11

Conclusion……….………………………………………………………………………………12

References……….………………………………………………………………………………13

Appendix………………………………………………………………………………………...14

1
ABSTRACT

An experiment is conducted in order to successfully calibrate a U/80 transducer which is

used in order to measure displacement by converting it into electrical energy. The experiment is
conducted by rotating a disc-shaft assembly, and reading the values from dial gauge, while
transducer varies as the position of magnets change, resistance of transucers changes as well. The
experiment is repeated for 0.05 mm steps. After the experiment is conducted, it is found that up
and down cycles of measurements are significantly different and measurements can change up to
4% for accuracy and up to 5% for precision.

2
INTRODUCTION

Since testing instruments may deteriorate in accuracy over time, it is really essential to perform
calibrations before testing or conducting experiments. Erroneous results can be minimized by
obtaining successful calibrations, therefore, the relation of inputs and outputs can be achieved more
reliably. Although testing instruments are manufactured in a good precision and certificated by the
manufacturer, it is always recommended to double check by conducting calibrations. Moreover,
these calibrations must be performed periodically due to degradations and deformations. Frequency
of use and environment are the main reasons for degradation. [1]

The purpose of this experiment was performing a successful calibration to U/80 transducer which
is used for displacement measurement and manufactured by Unimeasure Inc. A transducer is a
device that converts one form of energy to another form. U/80 transducer can be considered as a
sensor for this experiment since it converts displacement (mm) to resistance (ohm). In addition to
sensors, actuators are also a type of transducer which converts electrical signals to energy.

THEORY

Calibration methods are used to get better results in the experiment. In calibration, terms such as
accuracy, precision and hysteresis must be considered. In the calibration processes, comparisons
are made according to the standards of authorities in their fields (primary standards), to more
accurate values that have been tested (secondary standards) and to known input sources that have
been tested. [2]

In order to achieve more accurate results by providing calibration in the experiment, the transfer
function was created with the least square method and curve fitting.

Accuracy gives us the closeness between the measured value and the actual value. It is used to
reduce errors in individual measuring instruments with calibration. Individual measurements must
be calibrated individually to ensure the system's accuracy. The following equation (Eq. 1) should
be repeated for each individual measurement.

𝑑𝑒𝑣𝑖𝑎𝑡𝑖𝑜𝑛
𝐸𝑞. 1: 𝑎𝑐𝑐 = 𝑋 100
𝑜𝑢𝑡𝑝𝑢𝑡 𝑠𝑝𝑎𝑛

3
Precision is defined as the proximity of
each individual measurement to each
other. While calibration is used to
improve the accuracy of the experiment,
it does not changing the precision of the
experiment. The differences between Figure 1. Differences between precision and accuracy [3]
accuracy and precision is explained in
Figure 1.

Repeatability is one of the ways to calculate precision. Repeatability gives the variability in many
measurements made by the same person with the same instrument and technique. It is calculated
with the following equation (Eq. 2).

𝐸𝑞. 2: r = max value measured − min value measured

Linearity, used to state the performance of the measurement, defines the non-linear model of the
system. Eq.3 is used to calculate linearity. [4]

𝑅𝑎𝑣𝑔 − 𝑅𝑙−𝑠𝑞
𝐸𝑞. 3: 𝑙𝑖𝑛 = 𝑋 100
𝑜𝑢𝑡𝑝𝑢𝑡 𝑠𝑝𝑎𝑛

In the experiment, hysteresis is seen with the formation of a

characteristic loop in the input and output graph. Hysteresis
effect may occur due to factors affecting instruments such as
friction, magnetic and thermal effects, and deformations. The
Hysteresis effect is described in the measured variable versus
output reading graph in Figure 2. In Curve A, the changes in
the output reading are shown while the input increases
whereas in Curve B, the changes in the output reading while
the input decreases. If the output reading at different input Figure 2. Hysteresis effect
values is the same, the gap between these input values is Representation[5]

called a dead space. If there are two different output readings at the same input value that have
same absolute value, the absolute value of output reading is maximum output hysteresis.

4
The error originating from hysteresis is calculated with hysE (Eq. 4) and hys (%) (Eq. 5) given
below.

𝐸𝑞. 4 hysE = value approached from down direction − value approached from up direction

hysteresis error
𝐸𝑞. 5 hys (%) = 𝑋 100
𝑜𝑢𝑡𝑝𝑢𝑡 𝑠𝑝𝑎𝑛

Sensitivity, which indicates how the sensor is associated with unit change, gives a measure of the
sensitivity of the sensor to input. Sensitivity is calculated by using Equation 6 below.

𝑜𝑢𝑡𝑝𝑢𝑡 𝑠𝑝𝑎𝑛
𝐸𝑞. 6 s= 𝑋 100
𝑖𝑛𝑝𝑢𝑡 𝑠𝑝𝑎𝑛

Displacement transducer is designed to apply “Hall effect” principle to measure the displacement.
If a semiconductor armature which is connected to a circuit is exposed to a magnetic field, there
exists an electron concentration at one side of this armature and concentration of positive charge
on the other side. If a multimeter is connected to these sides, there will be a value in mV that can
be read from this multimeter. This voltage is called Hall voltage. However, if length of this
armature is much smaller than its width, this voltage will be a value that very close to zero.
Therefore, the path that electrons must travel gets longer and, consequently, resistance increases.
This phenomenon is called “magneto resistance effect”. [6]

SETUP

As indicated before, this experiment’s goal was to

calibrate the U/80 transducer and to analyze the results
acquired. The experiment setup consists of a digital
multimeter, a dial gauge, an U/80 displacement
transducer and a shaft-disc fixture (see in Figure 3).

By rotating the disc-shaft assembly, values are read

from dial gauge and transducer varies. The reason we
read different values from dial gauge is that there
exists a displacement sensor connected
Figure 3. Experimental setup.

5
 Figure 4. Dial gauge and displacement sensor. Figure 5. Magnets.
to this dial gauge (see figure 4). For the transducer, changing the position of the magnets (see figure
5)

causes a change in the flux density (3000 to 12000 gauss), therefore, the resistance value read from
multimeter varies. The resistance value is minimum at 3000 gauss and maximum at 12000 gauss.

First, we measured the resistance value at 1.3 mm read from the dial. Then we continued to the
measurements by increasing the displacement 0.05 mm steps until 1.9 mm and saved the data. At
the level of 1.9 mm, we exceeded that level a little bit then recorded the value by coming back to
same level so that we were able to see the hysteresis.

Afterwards, we started to decrease the displacement up to 0.7 mm and recorded the values in 0.05
mm steps. We applied the same method at the level of 0.7 mm. First, we reduced the displacement
to a value smaller than 0.7mm. Then, we increased the displacement to 0.7 mm and recorded the
value. The displacement was increased until 1.3 mm and value was recorded. These operations
were repeated 2 more time to obtain 3 cycles of input and output data. These data were used in
results and analysis of results sections.

RESULTS

According to experiment data, we have two main measurements which are displacement and
resistance from the transducer. The displacement range is [0.70,1.90] with 0,05 mm increments
and the resistance range is [229,433] as output.

6
Table 1. All the available experiment data collected and applied

Data of the experiment is presented on the Table 1. According to Table 1, the graphs of each cycle
is drawn and the hysteresis curves of up and down approaches are represented on the graphs.
Besides hysteresis, the least square line fit of the data is approximated and represented on the graphs
figure 1. Also, the average cycle is graphed in Figure 5 according to data from the Table 1.

Least square line is derived from the

Least Square Line
Excel using the average values that
450,00
can be seen from the Table 1 and the y = 165,94x + 114,09
400,00
equation of the LSL is taken from
Output(Ohm)

Excel directly. 350,00

After the initial drawing of the LSL, 300,00

all the remaining data has been

250,00
structured using Excel and
200,00
MATLAB.
0,50 1,00 1,50 2,00
Input(mm)

Figure 6. Least square line and its equation

7
Figure 7. Cycle 1 of experiment Figure 8. Cycle 2 of experiment

Figure 8. Cycle 3 of experiment Figure 9. Average Cycle of the Experiment

8
 Table 2. Repeatability In the Table 2, the repeatability and the hysteresis values are
Hysteresis
Displacement (Ohm) calculated
(mm) Repeatability (Ohm) and represented.
0,7 4 1
0,75 6 1 The accuracy, precision and terminal linearity calculations were
0,8 6 1
done on Excel and can be seen on the deviation plots below. The
0,85 6 1
0,9 6 1 process of obtaining these results are further explained on the
0,95 6 1
upcoming “Analysis of Results” chapter.
1 5 1
1,05 7 1
1,1 7 1
1,15 6 1
1,2 6 1
1,25 7 2
1,3 6 1
1,35 7 1
1,4 7 1
1,45 6 1
1,5 6 1
1,55 7 1
1,6 6 1
1,65 6 1
1,7 6 1
1,75 6 2
1,8 6 1
1,85 6 1
1,9 4 2

Figure 9. Deviation plot for accuracy

9
Figure 10. Deviation plot for precision

Figure 11. Terminal Linearity Graph

10
ANALYSIS OF RESULTS

With the data in hand, the results were obtained using the several different calibration specifics and
mostly the results were obtained at excel.

Table 3. Results

At first the LSL was fitted to the data and the line equation was derived. The
equation was a linear equation as the ax+b format was the most suited for
this type of linear relationship between input and output.

The least square line equation is 165.94x+114.09. According the this

equation and average values of the cycles the deviation plots were produced
and percent positive and negative accuracy and precision was acquired.
These results can be seen with 3 significant figures at table 3 next to this
writing.

As it can be seen from the values the positive accuracy limit is higher than negative accuracy limit
whereas positive precision limit is lower than negative precision limit in absolute values.

The accuracy is provided using this equation:

maximum deviation from LSL

Accuracy = x 100
output range

The precision is provided using this equation:

maximum deviation from Average

Precision = x 100
output range

Terminal linearity is provided using a similar equation but the deviation measured is from a line
that goes from the first average input to last average input.

The meaning of all these results shows us that the up and down cycles of measurements are
significantly different and accuracy and precision values for these are also different for sensitive
measurements changing up to 4 percent for accuracy and up to 5 percent for precision. Terminal
linearity is well within the accepted range when it is included that the system is both electronical
and mechanical meaning that lower terminal linearity is most likely not possible for this sensitivity
level of the equipment.

11
Sensitivity [7]: The least square line is based on the domain of Resistance (output) on [230,429]
and the displacement (input) on [0.70,1.90]. Therefore, the sensitivity of transducer is calculated
by Equation 6 as:

∆output ∆Resistance 429,376 − 230,248 198.90 Ω

s= = = = = 165.75
∆input ∆x 1,90 − 0,70 1,20 mm
When the system has higher sensitivity, we can get more accurate results from the system. We find
out our sensitivity is 165.75 Ohm/mm. It is enough to get necessary divisions on the output signal.

Hysteresis [8]: As we know that the hysteresis states the history dependence of systems. In basic,
it can be explained like the past results impact the new ones. So, in our experiment the calculations
are based on the up and down outputs. It represents the occurrence of hysteresis on our
measurements. We present the hysteresis as a percentage of output range as determined from the
least square line and calculated it by using Equation 5:

%hysteresis = %3,52
The repeatability values are represented on Table 2. As we can see from the table, the maximum
value of repeatability is 2 ohms. When we consider that the maximum value is small compared to
span of the sensor, the error of it can be seemed negligible than the others.

CONCLUSION

This experiment was conducted in order to successfully calibrate a U/80 transducer. Transducers
are used to transform displacement into electrical energy and thus making it able to measure easier.
Displacement transducer is designed to apply “Hall effect” principle to measure the displacement
as it uses the electrical charges on two sides. In order to understand if a transducer is successfully
calibrated; accuracy, precision and hysteresis are considered. In the experiment, a disc-shaft
assembly is rotated and values are read from dial gauge, while transducer varies as the position of
magnets change, resistance of transucers changes as well. The experiment is repeated for 0.05 mm
steps. Analyzing the results by drawing a best-fit line it’s been found that positive accuracy limit
is higher than negative accuracy limit whereas positive precision limit is lower than negative
precision limit in absolute values with measurements can change up to 4% for accuracy and up to
5% for precision while hysteresis is 3.52%.

12
REFERENCES

1) Beckwith, T. G., Marangoni, R. D. & Lienhard, J. H. (2011). Mechanical Measurements

(6th ed.). Pearson.
2) Engineering, H. (2020, March 17). Primary vs Secondary Standards. Harwood Engineering
Company, Inc. https://www.harwoodeng.com/articles/primary-vs-secondary-standards/
3) DNA Software Inc. (2019, November 8). Quick Guide to the Precision and Accuracy of
results from qPCR CopyCount. DNA Software. https://www.dnasoftware.com/our-
products/copycount-qpcr-analysis-copynumber/precision-and-accuracy/
4) Linearity and Calibration Verification: Overview. (n.d.). Data Innovations. Retrieved
April 7, 2021, from
https://www.datainnovations.com/sites/default/files/EE_Help/EE12/Content/LIN/linearity
___overview.htm
5) A.S. Morris and R. Langari. Measurement and Instrumentation: Theory and Application.
Elsevier Science, 2012. isbn: 9780123819604.
6) M. Nałȩcz and Z. Warsza, "Hall effect transducers for measurement of mechanical
displacements", Solid-State Electronics, vol. 9, no. 5, pp. 485-495, 1966. Available:
10.1016/0038-1101(66)90162-6 [Accessed 7 April 2021].
7) “What's Hysteresis?", Lassp.cornell.edu, 2021. [Online]. Available:
http://www.lassp.cornell.edu/sethna/hysteresis/WhatIsHysteresis.html. [Accessed: 07-
Apr- 2021].
8) R. Hogan, "How to Perform a Repeatability Test 3: | isobudgets", isobudgets, 2021.
[Online]. Available: https://www.isobudgets.com/how-to-perform-a-repeatability-test-3/.
[Accessed: 07- Apr- 2021].

13
APPENDIX

Matlab code:
clear all
close all
clc

%Data Reading from Excel Sheet

input=xlsread('DataTable.xlsx','A:A');
up_1=xlsread('DataTable.xlsx','J:J');
down_1=xlsread('DataTable.xlsx','K:K');
up_2=xlsread('DataTable.xlsx','L:L');
down_2=xlsread('DataTable.xlsx','M:M');
up_3=xlsread('DataTable.xlsx','N:N');
down_3=xlsread('DataTable.xlsx','O:O');
lsl=xlsread('DataTable.xlsx','I:I');
avg_up=xlsread('DataTable.xlsx','Q:Q');
avg_down=xlsread('DataTable.xlsx','R:R');

%Plotting Hysteresis and Least Square Line Graphs

p1=plot(input,up_1,'b')
hold on
plot(input,down_1,'b')
hold on
p3=plot(input,lsl,'r')
legend([p1 p3],{'Cycle1','Least Square Line'},'Location','NorthWest')
xlabel('Displacement (mm)')
ylabel('Resistance (ohm)')
title('Hysteresis Curve of Cycle 1');
grid on

figure(2)
p4=plot(input,up_2,'b')
hold on
plot(input,down_2,'b')
hold on
p5=plot(input,lsl,'r')
legend([p4 p5],{'Cycle2','Least Square Line'},'Location','NorthWest')
xlabel('Displacement (mm)')
ylabel('Resistance (ohm)')
title('Hysteresis Curve of Cycle 2');
grid on

figure(3)
p6=plot(input,up_3,'b')
hold on

14
plot(input,down_3,'b')
hold on
p7=plot(input,lsl,'r')
legend([p6 p7],{'Cycle3','Least Square Line'},'Location','NorthWest')
xlabel('Displacement (mm)')
ylabel('Resistance (ohm)')
title('Hysteresis Curve of Cycle 3');
grid on

figure(4)
p8=plot(input,avg_up,'b')
hold on
plot(input,avg_down,'b')
hold on
p9=plot(input,lsl,'r')
legend([p8 p9],{'Average Cycle','Least Square Line'},'Location','NorthWest')
xlabel('Displacement (mm)')
ylabel('Resistance (ohm)')
title('Hysteresis of Average Cycle');
grid on

15