Automatic Battery Monitoring System by Measuring Internal

Resistance Technique

by

Pongsakorn Sangpoem

A thesis submitted in partial fulfillment of the requirements for the

degree of Master of Engineering in
Microelectronics and Embedded Systems

Examination Committee: Dr. Mongkol Ekpanyapong (Chairperson)

Dr. Matthew N. Dailey
Assoc. Prof. Erik L.J Bohez

Nationality: Thai
Previous Degree: Bachelor of Engineering in Electronics Engineering
King Mongkut’s Institute of Technology Ladkrabang
Thailand

Scholarship Donor: Royal Thai Government Fellowship

Asian Institute of Technology

School of Engineering and Technology
Thailand
May 2018
 ACKNOWLEDGEMENTS

This is my opportunity to express my gratitude to persons who are involved in my

success of study at Asian Institute of Technology. Firstly, I would like to thank Dr.
Mongkol Ekpanyapong who is my advisor as well as Dr. Matthew N. Dailey and Assoc.
Prof. Erik L.J. Bohez who are my committee. They give me a lot of knowledge and
suggestion. Next, I am very grateful to Asst. Prof. Polphadung Phadungkul who shares
his experience and suggest to me. Finally, I am thankful for Ms. Pornpun Pugsawade,
Ms. Parichart Rodtong and Ms. Chowaret Sudsawaeng who are administrator in ISE
department.

ii
 ABSTRACT

VRLA battery is popular energy storage in power plant in Thailand. Battery cells will
be connected as series to increase the voltage level and connect to the charger.
However, when batteries are used for long time, they can deteriorate. Deterioration of
battery can lead to increasing in internal resistance and losing its capacity. To increase
the lifetime of battery, batteries have to be recovery when their internal resistance
change. At a result, battery monitoring system is very necessary of battery station. This
research will provide the study of methodology for monitoring battery by measurement
internal resistance technique. The battery monitoring system in this research is designed
to operate automatically and on-line station.

Keywords: VRLA battery, Battery monitoring, Battery measurement, Internal

resistance, On-line station

iii
 TABLE OF CONTENTS

CHAPTER TITLE PAGE

TITLE PAGE i
ACKNOWLEDGEMENTS ii
ABSTRACT iii
TABLE OF CONTENTS iv
LIST OF FIGURES v
LIST OF TABLES vi
LIST OF ABBREVATIONS vii

1 INTRODUCTION 1
1.1 Background 1
1.2 Problem Statement 1
1.3 Objectives 2
1.4 Limitations and scope 2
1.5 Research outline 2

2 LITERATURE REVIEW 3
2.1 Introduction of battery 3
2.2 Battery impedance measurement 3
2.3 Ohmic Measurement : The History and The Facts 9
2.4 Summary of internal resistance measuring method 10

3 METHODOLOGY 11
3.1 System Overview 11
3.2 System Design 11
3.3 Internal resistance measurement module designed 12
3.4 Battery selector module 20
3.5 Embedded system controller and software designed 21

4 EXPERIMENTS AND RESULTS 23

4.1 Simulation and Results of Hardware Designed 23
4.2 Experiment and Results of Hardware 26

5 CONCLUSION AND RECOMMENDATIONS 33

5.1 Conclusion 33
5.2 Recommendations 33

REFERENCES 34

iv
 LIST OF FIGURES (Cont'd)

FIGURE TITLE PAGE

Figure 2.1 DC method test setup 3

Figure 2.2 DC method test results 4
Figure 2.3 AC method simulation circuit 4
Figure 2.4 Voltage and current waveforms on battery in AC simulation 5
method
Figure 2.5 Circuit model of DC discharging of battery 5
Figure 2.6 DC discharge load circuit 6
Figure 2.7 DC discharge load device 6
Figure 2.8 AC impedance model of battery 6
Figure 2.9 Circuit model of AC impedance detection 7
Figure 2.10 Diagram of on-line battery impedance measurement system 7
Figure 2.11 The equivalent circuit of batteries 8
Figure 2.12 Diagram of internal resistance measurement module 8
Figure 2.13 Simplified battery equivalent circuit 9
Figure 3.1 System design of battery monitoring 11
Figure 3.2 Diagram of internal resistance measurement module 12
Figure 3.3 DC load method 12
Figure 3.4 AC conductance method 13
Figure 3.5 AC conductance load circuit 13
Figure 3.6 Low pass filter circuit 13
Figure 3.7 Signal result of AC current load circuit 14
(a) Voltage change of battery
(b) Current flow on 1 Ohm resistance
Figure 3.8 PWM sine wave generator 15
Figure 3.9 PWM signal output form Arduino 15
Figure 3.10 2nd-order low pass filter 16
Figure 3.11 Sine wave after filter the carrier frequency 17
(a) Sine wave 50 Hz
(b) Sine wave 100 Hz
(c) Sine wave 200 Hz
Figure 3.12 Passive high pass filter 18
Figure 3.13 Amplifier and active low pass filter 18
Figure 3.14 Zero crossing detector circuit 19
Figure 3.15 Input signal and output signal relation 19
Figure 3.16 Differential amplifier 20
Figure 3.17 Diagram of battery selector module 20
Figure 3.18 Embedded system and controller interfacing 21
Figure 3.19 Flowchart of algorithm of battery selector controller 21

v
 LIST OF FIGURES

FIGURE TITLE PAGE

Figure 3.20 Flowchart of algorithm of measurement module 22

Figure 4.1 Simulation circuit of AC load current and battery model 23
Figure 4.2 Simulation results of changing polarization resistance 24
(a) Smaller resistance
(b) Larger resistance
Figure 4.3 Simulation of changing electrochemical resistance 25
(a) Larger electrochemical resistance
(b) Smaller electrochemical resistance
Figure 4.4 Waveform of voltage and current of battery 26
Figure 4.5 Hardware for experiment 26
Figure 4.6 Results of parameters for first battery 27
(a) Conductance
(b) Battery voltage
(c) Temperature on terminal
Figure 4.7 Results of parameters for second battery 28
(a) Conductance
(b) Battery voltage
(c) Temperature on terminal
Figure 4.8 Results of parameters for third battery 29
(a) Conductance
(b) Battery voltage
(c) Temperature on terminal
Figure 4.9 Results of Midtronics for first battery 30
Figure 4.10 Results of Midtronics for second battery 30
Figure 4.11 Results of Midtronics for third battery 31
Figure 4.12 Relation graph of battery’s conductance from Midtronics and 32
hardware designed

vi
 LIST OF TABLES

TABLE TITLE PAGE

Table 2.1 Comparison of internal resistance estimation method 10

Table 4.1 Comparison of conductance measurement between hardware 31
designed and Midtronics

vii
 LIST OF ABBREVIATIONS

AIT Asian Institute of Technology

LC Language Center

viii
 CHAPTER 1

INTRODUCTION

1.1 Background

Power plant or power station is an industrial facility that is utilized for the generation
and distribution of electric energy. There are many types of power generation system
depending on the type of fuel used such as thermal power station, nuclear power station
and hydro-electric power station. Nowadays, most of the electric power can be
generated by using mechanical energy for propelling turbine and electricity generator.
Those mechanical energy as above is from converting other types of energy such as
thermal energy, water pressure energy and nuclear power. The Electric power will be
distributed to other power substation for dealing to users.

Electric energy is a basically essential factor for developing the country according to
acceptation that amount of electricity used can indicate economic growing of country.
Unfortunately, the process in power plant encounters problems frequently because
energy sources or machines wrong operate. Because of this issue, electric energy cannot
be generated and distributed to other substations. The way can solve this problem is
using the energy from storage to propel machines while technicians are repairing some
machines and solving the problems. Normally, batteries are used for storage the electric
energy in power plant.

Battery cells in power plant will be low voltage batteries but very high current. Each
cells of battery are serial connected, and the terminal of the end battery will be
connected to the battery charger. To expand lifetime of battery, the temperature in the
room which place batteries will be controlled properly. Moreover, measuring of
parameters such as terminal’s temperature, voltage, terminal resistance of battery will
be supplemented for diagnosis the quality of batteries.

1.2 Problem Statement

According to standard of battery in power plant system, batteries should have 20 years
of lifetime. However, lifetime of battery system in South Bangkok Power Plant of
Thailand is lower than 20 years. Some batches of battery have only 7 years of lifetime,
and are replaced because quality test is fail. Therefore, battery monitoring system is an
equipment which mounted on batteries for monitoring some parameters. However,
humans must control and monitor the system because current system is a stand-alone
system. Moreover, this system can provide only some parameters such as temperature
and voltage.

To reduce human’s roles, technical teams of power plant have to find other system or
equipment that can operate automatically. In addition, automatic system can measure
and update data regularly. Therefore, when some issues occur in battery systems, the
technical team can solve promptly that can increase the lifetime of battery.

1
1.3 Objectives

The elementary objective of this research is making the automatic battery monitoring
system by measuring internal resistance technique. The automatic battery monitor will
connect to terminal of each battery via wires for measure the parameter such as voltage,
conductance, ripple voltage, temperature. The battery monitor will be operated on
Raspberry Pi board. Moreover, Arduino board will be used as a buffer to improve the
performance of the system. Arduino board will read voltage and voltage change for
each battery and keep data for calling from main controller. The thermal sensors are
added for getting the temperature at terminals of batteries. All of parameters will be
processed and calculated on Raspberry Pi board. Finally, the complete data will be
collected and written in data log file.

1.4 Limitations and scope

This research study will cover the development of the automatic battery monitoring
system by measuring internal resistance technique which operated on Raspberry Pi
board. This system consists of two main parts. The first part is circuit board for testing
batteries. Another part is main controller board for collecting and calculating the data or
controlling some switch of circuit part. Because temperature has an effect some
parameters of batteries, we need to add calibration part in the system. Furthermore, for
high performance, batteries should be measured while they are uncharged.

1.5 Research outline

The outline of research proposal is shown as following.

In chapter 2, the literature review will be provided.

In chapter 3, the methodology of battery monitoring system will be provided.

In chapter 4, the experiment and results will be provided.

In chapter 5, the conclusion and recommendation will be provided.

2
 CHAPTER 2

LITERATURE REVIEW

2.1 Introduction of battery

Battery is electrochemical cells that used for storage of back-up electric energy. Inside
of battery contains two terminals and electrolyte solution. Battery can be classified as
two types that are flooded type and dry type. Additionally, dry type battery can be
classified as SLA (Seal Lead Acid) and VRLA (Valve Regulated Lead Acid). VRLA
battery is more secure than SLA battery because VRLA has valve for releasing
Hydrogen gases that are product from chemical reaction. However, both of VRLA and
SLA can be subdivided as Gel and AGM (Absorbed Glass Mat) battery.

2.2 Battery impedance measurement

Normally, internal impedance is usually used to indicate the deterioration of the battery
because the circuit of impedance measuring can be designed easily.

2.2.1 Stationary VRLA Battery Health Estimation by Resistance Measurement –

Comparison of DC and AC Test Methods.

a) DC test method

Figure 2.1 is the DC method test setup for estimating internal resistance in batteries.
This test method is used to estimate internal resistance of battery by discharging for a
few seconds with different testing conditions. 12 volt batteries are used in setup and
connected to switch and power supply. A bank resistor are connected and used as a load
test. A sensor board is used to measure and collect the voltage and current values.

Figure 2.1 DC method test setup

The result of testing is shown in figure 2.2. Batteries are discharged by 40 A and
repeated for 5 times

3
 Figure 2.2 DC method test results

b) AC test method

This test method is a simulation model to demonstrate AC method for estimation

internal resistance of battery. Constant current of 200 Hz sine wave signal is injected
into battery. The amplitude and phase of voltage are recorded and calculated as
impedance of battery. The simulation model is shown in figure 2.3.

Figure 2.3 AC method simulation circuit

4
The waveforms of the result are shown in figure 2.4. Current and voltage waveforms
with same frequency are recorded. The impedance of battery can be estimated by using
voltage and current value and phase shift of signal.

Figure 2.4 Voltage and current waveforms on battery in AC simulation method

2.2.2 Research On Integrated Method To Detect Internal Resistance Of Substation

Battery

This method will use DC discharging combined with AC impedance testing method to
detect the internal resistance of battery. DC discharge method can provide the accurate
and reliable result. However, the charger must be disconnected from the system. For AC
internal resistance detection method, the problem is tolerance of components and the
ripple current of the system. Combination of two methods can solve these problems.

a) DC discharge method

The circuit of DC discharge method for internal resistance detection of battery is shown
in figure 2.5. R, r, and K are represented external load resistor, internal resistor and
solid state relay respectively. The internal resistance can be calculated by equation (2.1).

Figure 2.5 Circuit model of DC discharging of battery

(2.1)

5
Overheating protection is inserted in DC discharge circuit as shown in figure 2.6. R2,
K1, and F are represented load resistor, normally open temperature switch and fuse
respectively. When relay K2 is closed state while K1 remains opened, discharge current
will flow through R2. When temperature of R2 reaches over limit, the switch K1 will be
closed to flow the current, and current that flows through R2 will be reduced. The DC
discharge load device for experiment is shown in figure 2.7.

Figure 2.6 DC discharge load circuit

Figure 2.7 DC discharge load device

b) AC impedance method

The AC impedance model of battery and circuit model of AC impedance detection are
shown in figure 2.8 and figure 2.9 respectively. Rm is represented polarization resistance
and activation polarization resistance, can go larger value if the battery deteriorates.

Figure 2.8 AC impedance model of battery

6
 Figure 2.9 Circuit model of AC impedance detection

Re is represented electrochemical resistance. AC impedance of battery can be calculated

by equation (2.2) as shown below.
(2.2)

2.2.3 On-line Battery Impedance Measurement

This impedance measurement of battery cells method in one string of battery cells can
be done by dividing each a string of battery into two portions a least. Each portion of
the string of battery will be imposed a loading current and given time for measurement.

Figure 2.10 Diagram of on-line battery impedance measurement system

Figure 2.10 is the diagram of on-line battery measurement system. Each battery cell of
each string portion will be measured by imposed loading current and given time. Then,
the connection will be changed to another string portion for providing a loading current

7
at given time to measure each battery cell of that string portion. The process will be
continuous for all of battery cells.

2.2.4 On-line Measurement of Internal Resistance of Lithium Ion Battery for

Electric Vehicle and its Application Research

The internal resistance of battery is comprised two parts that are ohmic resistance (Ro)
and polarization resistance (Rp). Ohmic resistance is formed from the material type of
the electrode in battery. As polarization resistance is formed from electrochemical
reaction. Additionally, inside of battery also includes the polarized capacitance (Cp).
The equivalent circuit of battery is shown in figure 2.11.

Figure 2.11 The equivalent circuit of batteries

On-line internal resistance measurement is included three parts are On-line

measurement module of internal resistance, USB-CAN and display of result (Host).
After internal resistance is measured, the data will be sent to computer via CAN bus
communication. The structure of inside of internal resistance measurement module is
shown in figure 2.12.

Figure 2.12 Diagram of internal resistance measurement module

8
2.3 Ohmic Measurement : The History and The Facts

Internal battery ohmic measurement is recognized that it is a measuring the internal

parameters of a battery cell that can be useful for determining a state of health of
battery. Moreover, these measurements have been accepted by the IEEE battery
working group and most battery manufacturers.

When a battery deteriorates and loses its capacity, the internal resistance will increases.
Albercorp experiences can present that the internal resistance of battery cell increases
by more than 25%, it will cannot store the charge by more than 80% of the rated
capacity.

Figure 2.13 Simplified battery equivalent circuit

Figure 2.13, this is a general battery equivalent circuit, which includes two resistors and
one capacitor in parallel with R2. The typical of the capacitor value is 1.5 farads per 100
ampere-hour of cell.

9
2.4 Summary of internal resistance measuring method

The basic concept of the methods for internal resistance of battery estimation is
summarized in table 2.1. For this research study, AC test method will be used for
estimation the internal resistance of battery because this technique can provide the
complex impedance.

Table 2.1 Comparison of internal resistance estimation method

Method Technique Advantages Disadvantages

Stationary VRLA DC test method - No noise in a - Cannot
Battery Health - discharging load system. consider
Estimation by current for a few - Do not need signal internal
Resistance time and injection. capacitance.
Measurement – recording voltage. - Easy method.
Comparison of DC AC test method - Can consider an - Noise signal
and AC Test Methods - Injecting sine effect of capacitance. can make an
wave signal into effect on the
the battery and system.
recording
amplitude and
phase of V and I.
Research On DC discharge - Easy method. - Thermal effect
Integrated Method To method - Accurate and of load resistor.
Detect Internal - Using load reliable.
Resistance Of resistor and - No noise influent.
Substation Battery recording V when
no load and
connect to load.
AC impedance - Can consider both - Circuit for AC
method of resistance and signal is more
- Using circuit to capacitance inside of complicated.
generate AC battery.
signal and
recording V, I and
phase.

10
 CHAPTER 3

METHODOLOGY

3.1 System Overview

This research study will be focused on three main parts. The first part is internal
resistance measurement module of battery for each cell by AC method. Temperature of
the battery’s terminal and voltage per cell of battery are necessary function of the
module, so a thermal sensor will be integrated on the module board. The second part is
battery selector module board that used to control the connection path for each battery
cell, and generate the sine wave. Another part is embedded system board used to control
all modules, collect data from measurement module and write data results in the log file.

3.2 System Design

Measurement module
Board
Embedded
System

Switch Board

Figure 3.1 System design of battery monitoring

The structure of the system design of battery monitoring is shown in figure 3.1. The
system will include internal resistance measurement module, controller board and
switch module. The controller will send the command of battery selection and all
parameters to switch board to select battery. After receiving command and parameters
from controller, the controller in switch module will process the switch to select the
battery and set parameters for measuring following command. Then, the measurement
module board will measure all parameters of battery by interface with switch module
and send result to embedded system board. The result will be calculated and collected in
data log file by respectively. The switch will change connection and path to select other
batteries for measurement. The process will continue until all of batteries are measured
and collected the data.

11
3.3 Internal resistance measurement module designed

Figure 3.2 Diagram of internal resistance measurement module

According to figure 3.2, the internal resistance measurement module is used to measure
several parameters in battery such as voltage per cell, internal resistance and
temperature of battery’s terminals. For the procedure of module, the load will be
connected to the battery and conduct current by condition of method. Voltage and
current change of the load and battery will be filtered noise out before amplifying.
Analog to digital converter will sampling the signal and send data to buffer for
collecting. Arduino board will be used as a buffer to collect the value before transferring
data to the main controller board via serial port. For thermal sensors, they will be
mounted on terminals of batteries and will be connected the module by switch module
board.

The method for testing internal resistance of battery has many methods such as taking
the resistance load or current injection.

3.3.1 The method of load current design

a) DC load method

For DC load method, the purpose of testing batteries is measuring pure ohmic. Firstly,
the current source will provide current and voltage for 2-3 seconds at 25-70 amperes.
The internal resistance can be calculated by using drop voltage and current. The
equivalent model of internal resistance of battery for DC load method is shown in figure
3.3.

Figure 3.3 DC load method

12
b) AC conductance method

For AC conductance method, the AC current at 80-100 Hz will be injected into

batteries. This range of frequency is mostly appropriate for inductive and capacitive
reactance converges. The equivalent model for internal resistance of battery estimation
can be shown in figure 3.4.

Figure 3.4 AC conductance method

3.3.2 AC conductance load design

AC current load circuit

Battery equivalent model
Figure 3.5 AC conductance load circuit

Figure 3.5 is the principle of AC conductance load is taking the AC current from the
battery. MOSFET is used to make a current source by connecting to 1 Ohm resistance at
source pin. 1 Ohm resistance should be low tolerance because it will be used to
calculate the flow current on this path. Operational amplifier is used to create the
voltage on Gate pin of MOSFET by connecting the reference voltage to non-inverting
pin of Op-Amp and connecting the voltage drop on 1 Ohm resistance to inverting pin as
a negative feedback. To eliminate noise on the system, low pass filter is added in
negative feedback by R6 5.6 KOhm and C3 10 nF that provide the cutoff frequency at
2.84 KHz is shown in figure 3.6.

Figure 3.6 Low pass filter circuit

13
 = (3.1)
= (3.2)
= 2.84 KHz (3.3)

100 Ohm (R4) is added to the Gate pin of MOSFET to protect the surge current from
charging effect. The figure 3.7 is represented the signal from the AC current load circuit
as the first graph waveform is voltage of battery and another waveform is the current
flowing on 1 Ohm resistance.

(a)

(b)
Figure 3.7 Signal result of AC current load circuit (a) Voltage change of battery
(b) Current flow on 1 Ohm resistance

14
3.3.3 PWM sine wave generator

Pulse Width Modulation Sine wave generator is a popular method to generate sine wave
signal by using microcontroller for higher accuracy. The voltage level of sine wave will
be compared with triangle signal base frequency to get the pulse width of each period of
square wave output that shown in figure 3.8.

Figure 3.8 PWM sine wave generator

Arduino board is used as a microcontroller for PWM sine wave generator to generate
the PWM output. Voltage comparing of sine wave and base frequency and pulse width
of output signal are processed in Arduino. The PWM output signal from Arduino is
shown in figure 3.9.

Figure 3.9 PWM signal output from Arduino

15
Output signal from Arduino must be filter out the carrier frequency by adding RC
passive low pass filter. R1, R2, C1 and C2 are connected as a 2nd-order low pass filter.
The circuit of 2nd-order low pass filter is shown in figure 3.10, and the sine wave signal
after filter carrier frequency is shown in figure 3.11.

Figure 3.10 2nd-order low pass filter

= (3.4)
√
= (3.5)
√
= 3558.81 Hz (3.6)

However, actual cutoff frequency of 2nd-order low pass filter is shown below.

= √ (3.7)

= 3558 √ (3.8)

= 2290.43 Hz (3.9)

16
 (a)

(b)

(c)
Figure 3.11 Sine wave after filter the carrier frequency (a) Sine wave 50 Hz
(b) Sine wave 100 Hz (c) Sine wave 200 Hz

17
3.3.4 Filter

Because AC current flow from battery, the AC voltage signal will occur on battery. The
AC voltage on battery will be filter by high pass filter because this signal includes DC
voltage level and AC voltage that cannot be amplified. High pass filter must be inserted
before voltage amplifier stage to prevent saturation due to DC offset voltage. R1, R2
and C1 are used to be a high pass filter at 0.67 Hz cutoff frequency, Moreover, R1 and
R2 can be used to create the voltage reference of input of operational amplifier. The
high pass filter circuit is shown in figure 3.12

Figure 3.12 Passive high pass filter

= (3.10)

= (3.11)
= 0.67 Hz (3.12)

3.3.5 Voltage amplifier

After high pass filter stage, amplifier will amplify the voltage signal because the signal
from battery is very small. Voltage gain of amplifier should proper for signal, if gain is
too high, the signal will be saturation. LM2904 IC is used to build a voltage amplifier
by connect R3 and R4 as a negative feedback to control voltage gain at 101 V/V.
However, the amplifier should be control the band width to prevent the noise at high
frequency by add C2 parallel with R3. The voltage amplifier and active low pass filter is
shown in figure 3.13.
High pass filter Amplifier and low pass filter

Figure 3.13 Amplifier and active low pass filter

18
3.3.6 Zero crossing detector

Zero crossing detector is used to detect sine wave signal when the signal is higher than
zero voltage or ground of AC signal. For this research study, zero crossing detector can
help controller board to collect the highest peak or lowest peak voltage, thus the
program will not be complicate. The output of zero crossing detector will be connected
to interrupt pin of Arduino board. R1, R2, R3 and R4 are used to make voltage
reference Vcc/2, and C1 used to be AC coupling capacitor. However, all resistors and
capacitors must be proper with the frequency of signal. The circuit of zero crossing
detector is shown in figure 3.14, and the signal input and output relation is shown in
figure 3.15.

Figure 3.14 Zero crossing detector circuit

vin

Vout

Figure 3.15 Input signal and output signal relation

3.3.7 Differential amplifier

Since parasitic resistance in wires can have an effect in signal, it can generate offset
voltage signal and make the error result. LM2904 used to build differential amplifier by
connect 10Kohm resistors to make gain 1 V/V. The output voltage equals to V2 – V1.
The differential amplifier circuit is in figure 3.16.

19
 Figure 3.16 Differential amplifier

3.4 Battery selector module

To make the system can monitor multiple batteries, the battery selector module is added
in the system. Arduino board is used as a controller to provide the logic control for
multiplexer. Multiplexer will decode the logic from Arduino and send to trig the relays
or analog switches. Switch channel which trigged by multiplexer will connect to battery
and measurement module. The block diagram of battery selector module is shown in
figure 3.17.

Figure 3.17 Diagram of battery selector module

20
3.5 Embedded system controller and software designed

3.5.1 Embedded system designed

For this research study, the system is operated on embedded system board which
communicates to two microcontroller boards in measurement module and battery
selector module via USB serial port. Raspberry Pi is used as an embedded board will
send the command and parameters to Arduino in battery selector module, and receive
data from Arduino in measurement module board. I/O interface is used to communicate
between two Aduino board. The embedded system structure is shown in figure 3.18.

Figure 3.18 Embedded system and controller interfacing

3.5.2 Software designed for battery selector module

Figure 3.19 Flowchart of algorithm of battery selector controller

21
Figure 3.19 is the algorithm of controller in selector module board. Firstly, the Arduino
board will receive the command and parameters such as voltage, frequency, battery
channel, and process PWM from the command. When finishing to process PWM,
Arduino will send status to measurement board, and wait feedback status to continue for
releasing PWM signal. Finally, Arduino will stop releasing PWM if the status of
measurement module is changed.

3.5.3 Software designed for measurement module

Figure 3.20 is shown the flowchart of algorithm for measurement module. Controller in
measurement module will repeat reading ready status from selector module. If status of
battery selector module is ready, Arduino in measurement module will read the voltage
level of battery before sending ready status back to selector module. For stable signal,
microcontroller must add some delay before reading the voltage and current from
LT2440 ADC via SPI. After finishing to reading data, the results will be send to
Raspberry Pi board by USB port. Finally, Arduino on measurement module will set
status to zero for waiting next measurement.

Figure 3.20 Flowchart of algorithm of measurement module

22
 CHAPTER 4

EXPERIMENTS AND RESULTS

In this chapter, the experiments will explain the procedures for testing the designed
equipment which the methods to design are discussed in previous chapter. Experiments
will focus on the main objective of this research study that is monitoring internal
resistance or internal conductance. The experiments and results consist of simulation of
hardware designed and hardware designed.

4.1 Simulation and Results of Hardware Designed

4.1.1 AC load current circuit designed

The major circuit designed of the automatic battery monitoring system by measuring
internal resistance technique is the AC load current circuit for sinking the current and
measuring voltage changed. The simulation circuit is designed as two main parts which
include the AC current sink and the equivalent model of battery by specify the
parameters in equivalent model of battery according to the paper about battery. 50 Hz
and 1 Volt sinusoidal wave is injected to the system and voltage changed of battery will
be measured. The simulation circuit of AC current load for hardware designed is shown
in figure 4.1.

Figure 4.1 Simulation circuit of AC load current and battery model

a) The effect of polarization resistance

The effect of polarization resistance (Rp) will be find by changing the values and
monitoring voltage changed of battery. Polarization resistance can increase if battery is
deterioration. In simulation, Rp will be chaged from 100 µohm to 1000 µOhm which
100 µOhm is a value of known good battery. The volate changed of battery and phase
shift of current and voltage will be monitored. The waveform results of signal are
shown in figure 4.2.

23
 (a)

(b)
Figure 4.2 Simulation results of changing polarization resistance
(a) Smaller resistance (b) Larger resistance

b) The effect of electrochemical resistance

The electrochemical resistance (Re) is a resistance that parallel with internal capacitance
in equivalent model of battery. The electrochemical resistance does not change when
batteries are deterioration. However, Re can be used to explain the phase shift of
voltage and current. In simulation circuit, the Re is decreased from 100 µOhm to 1
µOhm, and the phase of voltage and current will be focused. The result of simulation is
shown in figure 4.3.

24
 (a)

(b)
Figure 4.3 Simulation of changing electrochemical resistance
(a) Larger electrochemical resistance (b) Smaller electrochemical resistance

4.1.2 Discussion of simulation

According to simulation, the result of effect of polarization resistance and

electrochemical resistance can clarify that the small electrochemical resistance which
are very small compare to polarization resistance and does not change, will not has an
effect in phase shift of voltage and current signal. Moreover, when polarization
resistance is bigger, the phase shift will decrease. To prove the simulation, simulation
circuit is applied to build the real circuit and measure the signal. The waveforms are
shown in figure 4.4 which including voltage (yellow graph) and current (blue graph).
25
 Figure 4.4 Waveform of voltage and current of battery

4.2 Experiment and Results of Hardware

4.2.1 Hardware experiment and results

Figure 4.5 is the complete hardware which used to do an experiment. The procedures of
the experiment starts by connect the wires to plus and minus terminals on battery and
mount the temperature sensor on each terminal. When the software in embedded system
board work, the parameters for testing are required such as measurement channel,
current load test, frequency for test and quantity of repeatability. For this experiment,
the parameters are used including channel 1, current 1 Ampere peak to peak, frequency
50 Hz and 50 times of repeatability.

Figure 4.5 Hardware for experiment

26
a) Results for first battery

Figure 4.6 is the graphs of measuring result for first battery cell. The standard deviation
of conductance is 1.658743 Siemen and battery voltage is 0.0037 Volt which show that
the measurement has high precision. However, the standard deviation of temperature on
terminals depends on the ambient temperature.

(a)

(b)

(c)
Figure 4.6 Results of parameters for first battery (a) Conductance
(b) Battery voltage (c) Temperature on terminal
27
(b) Results of second battery

Figure 4.7 is the results of measurement for second battery cell. The standard deviation
of conductance is 20.14981 Siemen and voltage of battery is 0.002031 Volt.

(a)

(b)

(c)
Figure 4.7 Results of parameters for second battery (a) Conductance
(b) Battery voltage (c) Temperature on terminal

28
c) Results of third battery

Figure 4.8 is shown the results of measurement for third battery cell. The standard
deviation of conductance for this battery cell is 10.46463 Siemen, and battery voltage is
0.000842 Volt.

(a)

(b)

(c)
Figure 4.8 Results of parameters for third battery (a) Conductance
(b) Battery voltage (c) Temperature on terminal
29
4.2.2 Comparison of hardware experiment results and Midtronics results

The results of conductance measurement by using hardware in research study will be

compared with the result of conductance from Midtronics’s meter. Midtronics is a
commercial company which has developed technologies for battery management. The
accuracy of hardware designed for this research will be calculated comparing to
Midtronics tool.

a) Comparison result of first battery

For this battery cell, the conductance which is measured by Midtronics is 349 Siemen
while the hardware can measure the value around 282 Siemen. Therefore, the accuracy
is 80.80%. Figure 4.9 is the measuring results on Midtronics screen.

Figure 4.9 Result of Midtronics for first battery

b) Comparison result of second battery

The conductance from Midtronics is 1505 Siemen, and the conductance from hardware
in this research is 1461 Siemen. Thus, the accuracy is 97.08%. The measuring results
are shown on screen as figure 4.10

Figure 4.10 Result of Midtronics for second battery

30
c) Comparison result of third battery

The conductance of battery from Midtronics is 1001 Siemen while the conductance
measured by hardware is 963 Siemen. The accuracy which is compared with Midtronics
is 96.20%. The results of measurement from Midtronics is shown in figure 4.11.

Figure 4.11 Result of Midtronics for third battery

d) Summary of comparison between the hardware designed and Midtronics

Table 4.1 is summary of conductance value which measured by hardware designed

compared with value from Midtronics. Conductance values of three battery cells
measured by hardware of this research are lower than value from Midtronics. In the
table 4.1, the errors of measurement are summarized in term of errors and percentage of
relative error. However, the percentage of error or accuracy that low or high might not
indicate the real accuracy because it depends on raw value. For example, conductance
of cell 2 has an error by 44 and %error 2.92, while conductance of cell 3 has an error 38
but %error is 3.80 which more than cell 2. For the instrument which has good precision,
it can be calibrated by comparing with the reference tools for improving the accuracy.
In figure 4.12, this is a graph to calculate the offset value by finding the equation of
graph.

Table 4.1 Comparison of conductance measurement between hardware designed

and Midtronics

Battery Conductance (Siemen) Error

%Error %Accuracy
cell Hardware designed Midtronics (Siemen)
1 282 349 67 19.20 80.80
2 1461 1505 44 2.92 97.08
3 963 1001 38 3.80 96.20

31
Figure 4.12 Relation graph of battery’s conductance from Midtronics and
hardware designed

32
 CHAPTER 5

CONCLUSION AND RECOMMENDATIONS

5.1 Conclusion

In this research study, the automatic battery monitoring system was developed for
monitoring the quality and deterioration of battery cell in power plant. The technique
which used to design the hardware is internal resistance measurement by sinking the AC
current and measuring voltage changed for calculation the resistance following Ohm’s
law. The hardware was designed to measure all parameters of battery for 4 battery cells
per one hardware. However, this system used AC conductance method to find the
internal resistance of battery which is low level signal. The main problem is the system
will be sensitive with all frequency signals such as noise. Thus, the system must has
high sensitivity for interested signal, and compress the noise signals as much as
possible.

According to experiment, the experimental results of measuring conductance compared

with the results by Midtronics which is an equipment for internal resistance
measurement designed from commercial company in USA. The results usually have
high precision but the accuracies are different.

5.2 Recommendations

This research is proposed to design the methodology and technique to measure internal
resistance in battery. Although the system designed is the best and high performance,
the errors can occur due to several factors such as voltage supply which used in circuit,
the noise in natural and the probes pin that used to contact with poles of battery.

33
 REFERENCES

[1] Glenn Alber, Albercorp Boca Raton, FL (2003), OHMIC MEASUREMENTS:

THE HISTORY AND THE FACTS

[2] Feng Zhenghua, zhili, Wang Bin (2012). Research On Integrated Method To
Detect Internal Resistance Of Substation Battery

[3] ESwararao Kona (2016). Stationary VRLA Battery Health Estimation By

Resistance Measurement – Comparison of DC and AC Test Methods

[4] John W, Wurst, Chester Township, Morris County, N.J. (1992), ON-LINE
BATTERY IMPEDANCE MEASUREMENT

[5] H. Zhang (2014), On-line Measurement of Internal Resistance of Lithium Ion

Battery for Electric Vehicle and its Application Research

[6] EAST PENN Expertise and American Workmanship (2015), Valve-Regulated

Lead-Acid (VRLA): Gelled Electrolyte (Gel) and Absorbed Glass Mat (AGM)
Batteries

34