designing a food spoilage monitoring device that uses near field

ECE 4600 Group Design Project

Designing a Food Spoilage Monitoring Device thatuses Near Field Communication Techniques

byGroup 04

Adam DueckBozhou Du

Damilola OdunsiDarlington Ogundu

Shucheng Zheng

Final report submitted in partial satisfaction of the requirements for the degree of

Bachelor of Science in Electrical and Computer Engineering in the

Faculty of Engineering of the University of Manitoba

Academic Supervisor(s)

Dr. Gregory Bridges, Ph.D., P.Eng., Sharmistha Bhadra, M.Sc.Department of Electrical and Computer Engineering

University of Manitoba

Date of Submission

March 10, 2014

Copyright 2014 Adam Dueck, Bozhou Du, Damilola Odunsi, Darlington Ogundu,Shucheng Zheng

NFC Food Spoilage Monitoring Device

Abstract

Food spoilage has detrimental effects both in terms of health risks for consumers as well

as increased costs for producers. This prompted the development of a system for detecting

food spoilage that was compact, low-maintenance, and simple to manufacture. The design

of said system is discussed in this report. The above-mentioned requirements were met

by creating a battery-less, chip-less, NFC-based sensor capable of detecting gases that are

released by spoiling food through changes in pH.

The pH sensor can be broken down into two main components: a pair of electrodes

and an LC resonant circuit. The electrodes comprise a sensing electrode and a reference

electrode, both of which are coated with hydrogel. The resonant circuit contains a variable

capacitor and an inductive coil. The gases that are released by the decaying food are

absorbed by the hydrogel, causing a change in its pH. These changes in pH are marked by

the resulting change in the natural frequency of the resonant circuit. A method for wirelessly

measuring these changes in frequency using NFC was also developed. A measured change in

the resonant frequency can therefore be used to determine if a change in pH has occurred,

thereby marking food spoilage.

The values of the pH sensor electronic components were calculated and the circuit was

simulated in Multisim. The schematic was developed into a PCB design using Ultiboard,

and the pH sensor tag was manufactured. The sensor was tested by applying a DC voltage

in place of the electrodes, and a linear change of frequency with voltage was determined

using an impedance analyzer. The electrodes were then fabricated, and connected to the

tag, completing the pH sensor. Meanwhile the components for controlling and powering

the interrogator were designed, simulated in Multisim, and soldered to a prototype board.

During final testing, readings were taken with both the interrogator and an impedance

analyzer. We were unable to obtain the expected results with the interrogator. However,

the tests performed with the impedance analyzer showed a clear linear relationship between

the resonant frequency of the sensor and pH, proving that our NFC-based pH sensor works.

- i -


Contributions

Ad

amD

uec

k

Boz

hou

Du

Dam

ilol

aO

du

nsi

Dar

lin

gton

Ogu

nd

u

Shu

chen

gZ

hen

g

Fabrication of the Electrodes Testing of the Electrodes Design of Glass Chamber Design build and test hand wound coil Design and test printed spiral coil Design build and test resonant circuit with hand wound coil Design build and test resonant circuit with printed spiral coil Designing the Switch Controller (Interrogator) Building the Switch Controller (Interrogator) Designing the Data Acquisition System (Interrogator) Testing the Interrogator Setup

Legend: Lead task Contributed

- ii -


Acknowledgements

We would like to thank our supervisors, Dr. Bridges and Sharmistha Bhadra, for their

advice and support throughout the course of this project. We would also like to thank Tao

Chen for milling our PCB, Sinija Janjic for ordering our materials, Cory Smit for fabricating

our glass chamber and Daniel Card for helping us with our design.

- iii -

NFC Food Spoilage Monitoring Device TABLE OF CONTENTS

Table of Contents

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . i

Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii

List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x

1 Introduction 1

1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.2 Design Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.3 Specification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

1.4 Report Organisation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2 Background 4

2.1 Food Spoilage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

2.2 pH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2.3 Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2.4 NFC Technololgy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.4.1 NFC System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.5 Time Domain Gating Interrogation . . . . . . . . . . . . . . . . . . . . . . . 10

2.5.1 Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

2.5.2 Amplitude-Based Approach . . . . . . . . . . . . . . . . . . . . . . . 12

2.5.3 Frequency-Based Approach . . . . . . . . . . . . . . . . . . . . . . . 12

3 Design Methodology 13

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

3.2 Electrodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

- iv -


3.2.1 Fabrication of Solid-State Reference Electrode . . . . . . . . . . . . 14

3.2.2 Interfacing Electrodes to the Tag . . . . . . . . . . . . . . . . . . . . 14

3.3 NFC Tag . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

3.3.1 Varactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

3.3.2 Sensor Coil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

3.3.3 Resonant Circuit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3.4 Interrogator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

3.4.1 Portable Interrogator . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

3.4.2 Stationary Interrogator . . . . . . . . . . . . . . . . . . . . . . . . . 25

4 Testing Procedure and Results 37

4.1 Test Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

4.1.1 DC Generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

4.1.2 Impedance Analyzer . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

4.1.3 Function Generator . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

4.1.4 Oscilloscope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

4.1.5 Pulse Generator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

4.1.6 Digital Multimeter . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

4.2 Individual Component Testing . . . . . . . . . . . . . . . . . . . . . . . . . 39

4.2.1 Electrodes testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

4.2.2 Sensor Coil testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

4.2.3 NFC Tag Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

4.2.4 TDG Interrogator Coil Testing . . . . . . . . . . . . . . . . . . . . . 55

4.2.5 Testing of Interrogator Timings . . . . . . . . . . . . . . . . . . . . . 56

4.2.6 Switch Controller Testing . . . . . . . . . . . . . . . . . . . . . . . . 59

4.2.7 Data Acquisition System Testing . . . . . . . . . . . . . . . . . . . . 65

4.3 Final Design Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66

4.3.1 Interrogator and NFC Tag Testing . . . . . . . . . . . . . . . . . . . 66

4.3.2 pH Sensor and Impedance Analyzer Testing . . . . . . . . . . . . . . 70

5 Conclusions 76

References 78

Appendix A Appendix A 78

Appendix B Appendix B 79

- v -


Appendix C Curriculum Vitae 83

- vi -

NFC Food Spoilage Monitoring Device LIST OF FIGURES

List of Figures

2.1 The Logarithmic pH Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2.2 NFC Tag Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2.3 Time domain gating-based interrogator block diagram . . . . . . . . . . . . 11

2.4 Time domain gating timing diagram . . . . . . . . . . . . . . . . . . . . . . 11

3.1 Electrodes immersed in hydrogel . . . . . . . . . . . . . . . . . . . . . . . . 15

3.2 Varactor capacitance vs reverse bias voltage curve . . . . . . . . . . . . . . 16

3.3 Inductance value of the sensor coil measured by the impedance analyzer . . 17

3.4 Single layer planar square spiral coil . . . . . . . . . . . . . . . . . . . . . . 20

3.5 An equivalent circuit diagram of the PH sensor and coupled interrogator coil 21

3.6 Portable Interrogator Receiver - design 1 block diagram . . . . . . . . . . . 24

3.7 Portable Interrogator Receiver design 2 block diagram . . . . . . . . . . . . 25

3.8 Stationary interrogator setup block diagram . . . . . . . . . . . . . . . . . . 27

3.9 Simulation of 555 timers in astable configuration . . . . . . . . . . . . . . . 29

3.10 Simulation of 555 timers in monostable configuration . . . . . . . . . . . . . 30

3.11 Configuration of the Delay Chip . . . . . . . . . . . . . . . . . . . . . . . . 31

3.12 Simulation Results of the Delay Chip . . . . . . . . . . . . . . . . . . . . . . 32

3.13 A: 10 Hz sine wave with noise in time domain . . . . . . . . . . . . . . . . . 35

3.14 B: 10 Hz sine wave with noise in frequency domain . . . . . . . . . . . . . . 35

3.15 Schematic of the +5 V voltage regulator . . . . . . . . . . . . . . . . . . . . 36

3.16 Schematic of the -5 V voltage regulator . . . . . . . . . . . . . . . . . . . . 36

4.1 Complete test setup with multimeter attached (voltage shown is still changing) 40

4.2 pH level vs. voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

4.3 Setup of the testing of the hand wound air cored coil . . . . . . . . . . . . . 43

4.4 Hand wound air core coil . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

- vii -

NFC Food Spoilage Monitoring Device LIST OF FIGURES

4.5 Inductive value of the hand wound coil . . . . . . . . . . . . . . . . . . . . . 46

4.6 Self-resonant frequency of the hand wound coil . . . . . . . . . . . . . . . . 46

4.7 Milled PCB coil . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

4.8 Top view NFC tag testing set up . . . . . . . . . . . . . . . . . . . . . . . . 49

4.9 Front view NFC tag testing set up . . . . . . . . . . . . . . . . . . . . . . . 49

4.10 Top view of the resonant circuit . . . . . . . . . . . . . . . . . . . . . . . . . 51

4.11 Front View of the resonant circuit . . . . . . . . . . . . . . . . . . . . . . . 51

4.12 Resonant frequency when bias voltage is 0 V . . . . . . . . . . . . . . . . . 53

4.13 Resonant frequency when bias voltage is 0.4 V . . . . . . . . . . . . . . . . 53

4.14 Resonant frequency versus voltage applied to NFC tag . . . . . . . . . . . . 54

4.15 Largest operational distance for sensor coil and interrogator coil . . . . . . . 55

4.16 Testing substitute resonator with impedance analyzer . . . . . . . . . . . . 57

4.17 Input (blue) and output (pink) of the first flip flop . . . . . . . . . . . . . . 60

4.18 Input (blue) and output (pink) of the first 555 timer . . . . . . . . . . . . . 61

4.19 Input (pink) and output(blue) of the programmable delay chip . . . . . . . 62

4.20 Output 1 (pink) and output 2(blue) to the switches . . . . . . . . . . . . . . 63

4.21 Unit test of output 1 and a switch . . . . . . . . . . . . . . . . . . . . . . . 64

4.22 Unit test of output 2 and a switch . . . . . . . . . . . . . . . . . . . . . . . 65

4.23 Natural frequency obtained when voltage is 0 V . . . . . . . . . . . . . . . . 67

4.24 Natural frequency obtained when voltage is 0.1 V . . . . . . . . . . . . . . . 68

4.25 Natural frequency obtained when voltage is 0.2 V . . . . . . . . . . . . . . . 69

4.26 Estimated pH vs natural frequency . . . . . . . . . . . . . . . . . . . . . . . 70

4.27 Final test setup with the impedance Analyzer . . . . . . . . . . . . . . . . . 71

4.28 Test result with ammonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73

4.29 Test result with deionized water . . . . . . . . . . . . . . . . . . . . . . . . . 73

4.30 Test result with acetic acid . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

4.31 Plot of pH vs resonant frequency . . . . . . . . . . . . . . . . . . . . . . . . 75

A.1 The Built Switch Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

- viii -

NFC Food Spoilage Monitoring Device LIST OF TABLES

List of Tables

1.1 Device Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

3.1 Core Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

4.1 Voltage responses of the electrodes to varied pH levels . . . . . . . . . . . . 41

4.2 Dimensions of the Air Core Coils . . . . . . . . . . . . . . . . . . . . . . . . 44

4.3 Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

B.1 Budget . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

- ix -


Nomenclature

- x -


Abbreviations Description

ADC Analog to Digital.

Ag/AgCl Silver/Silver Chloride.

CW Continuous Wave.

DC Direct Current.

NFC Near Field Communication.

IC Integrated Circuit.

ISM Industrial, Scientific, Medical.

KCl Potassium Chloride.

LC Inductor-capacitor.

LCD Liquid-crystal Display.

M Molar Concentration.

MMO Mixed-metal Oxide.

NaCl Sodium Chloride.

NH4OH Ammonium Hydroxide.

PCB Printed Circuit Board.

pH Power of Hydrogen.

PVC Polyvinyl Chloride.

Q Factor Quality Factor.

RF Radio Frequency.

RFID Radio Frequency Identification.

TDG Time Domain Gating.

THF Tetrahydrofuran.

TMAO Trimethylamine-N-oxide.

TTL Transistor-transistor Logic.

VCXO Voltage Controlled Crystal Oscillator.

Vpp Peak-to-peak Voltage.

- xi -


Term Definition

Anode Electrode that is oxidized.

Bias Voltage An applied voltage used to set the operating point

of an electronic component.

Cathode Electrode that is reduced.

Cutoff Frequency The frequency that marks the division between the

pass band and the stop band of a filter. For a low

pass filter, frequencies less than the cutoff will be

passed while higher frequencies will be blocked.

Continuous Wave An electromagnetic wave of constant frequency

and amplitude.

Hydrogel A hydrophilic polymer; gel with a high water content.

ISM band Industrial, scientific and medical band; An

unregulated frequency band that is often used for

low-power, short-range communication technologies,

including NFC.

Resonant Frequency The frequency at which impedance is maximum

when plotting the frequency response of an resonator.

As Q increases, the resonant frequency approaches

the natural frequency.

Natural Frequency The frequency at which a resonator will tend to

oscillate under zero input conditions. As Q

increases, the resonant frequency approaches the

natural frequency.

Resonator A system that naturally oscillates at a certain

frequency with greater amplitude than other

frequencies.

Reduction When one chemical species gains electrons in a

chemical reaction.

LC Resonator A resonator that is composed of an inductor

and a capacitor.

Lipolysis The chemical breakdown of lipids.

Power of hydrogen A logarithmic scale of the concentration of

hydrogen ions in an aqueous solution.

- xii -


Term Definition

Time Domain Gating An approach to wireless interrogation;

transmit and receive operation are separated into

discrete phases, with a short time delay in between

the two.

Molar Concentration The concentration of a solution in moles of solute

per liter of solvent

Near Field Communication A form of communication based on magnetic

coupling between inductive coils

Oxidation When one chemical species loses electrons in a

chemical reaction.

Printed circuit board Acts as a platform for electronic components, with

copper traces acting as interconnects and a

non-conductive substrate providing mechanical

support.

Proteolysis The chemical breakdown of proteins.

Syneresis When liquid is extracted or expelled from a gel

Transistor-transistor Logic A type of digital circuit based on bipolar junction

transistors. Typical TTL circuits are powered with

a 5V supply, with logic low between 0V and 0.8V

and logic high between 2.2V and 5V.

Varactor A special purpose diode that acts as a variable

capacitor controlled by a DC reverse bias voltage.

- xiii -

NFC Food Spoilage Monitoring Device 1. Introduction

Chapter 1

Introduction

1.1 Motivation

Every year, food spoilage leads to wastage and millions of food poisoning cases worldwide.

The absence of adequate food monitoring technology could compromise the quality of food

products during transit. The ability to wirelessly monitor the quality of food produce

reduces costs for producers and ensures consumers health and safety hence making food

quality measurement an indispensable requirement. Current means of monitoring food

quality are through the use of sensors which observe changes in temperature, release of

gases (such as ammonia) and secretion of enzymes. These methods are not entirely practical

due to cost-constraints, bulky proportions, complex fabrication processes and the need for

regular maintenance. Due to the aforementioned reasons, we were motivated to develop a

simple and cost-effective wireless food monitoring device.

1.2 Design Overview

This device employs Near Field Communication (NFC) and comprises a pH sensor and an

external interrogator that interact to provide information on the quality of food products.

- 1 -

NFC Food Spoilage Monitoring Device 1.3 Specification

The pH sensor is further divided into electrodes and resonant circuit. The electrodes of

the sensor are coated with hydrogel; a hydrophilic polymer which acts as an electrolytic

solution where the electrodes interact. The resonant circuit consists of an inductive coil

and sensor electronics. The interrogator consists of an inductive coil, discrete components,

a switch controller and a data acquisition system. The interrogator uses a time-domain

gating method to interact with the pH sensor through inductive coupling.

In this project, the goal was to detect food spoilage through pH change. The change

in pH of the food being tested causes a change in the voltage across the electrodes and this

voltage change varies the capacitance of the NFC tag. The NFC tag will therefore resonate

at a different frequency as pH changes. The interrogator will obtain this new frequency

through inductive coupling which will be used to detect pH changes. This will enable the

user to determine whether or not food spoilage has occurred.

1.3 Specification

The table below shows the specifications of the entire food spoilage monitoring device. They

have been revised based on the results obtained from our final testing.

Table 1.1: Device Specifications

Parameter Value

Minimum Interrogation Distance 3 cm

Resonant Frequency 12.37 - 13.48 MHz

Resonance shift due to pH 1 MHz

pH Operating Range 2-12 pH

pH Sensing Accuracy 0.5 pH

Maximum Sensor Length 9 cm

Maximum Sensor Width 5 cm

Maximum Sensor Thickness 0.3 cm

- 2 -

NFC Food Spoilage Monitoring Device 1.4 Report Organisation

1.4 Report Organisation

This report is organised in order to provide the reader with an understanding of the design

and implementation of the food spoilage monitoring device. Chapter 2 is an overview of the

concepts used in our design and Chapter 3 provides reasons for our design choices. Chapter

4 deals with the testing procedure and results obtained from our design and Chapter 5

discusses the conclusions made from our project.

- 3 -

NFC Food Spoilage Monitoring Device 2. Background

Chapter 2

Background

2.1 Food Spoilage

Food spoilage is the process in which food deteriorates over time. It is characterised by any

change in the food product that renders it unattractive and unsuitable for consumption [1].

This process leads to food becoming inedible and sometimes poisonous. Spoilt food is unfit

for human consumption and if consumed, can lead to serious illness and sometimes death.

Due to the negative effects of food spoilage, there are a number of preservation techniques

in place to prevent food spoilage [1]. Traditional methods of preserving food like drying,

heating and fermentation are aimed at deactivating the spoiling microorganisms and they

work to an extent because microbial spoilage is the most common cause of food spoilage [1].

There are various specific causes of food deterioration and they differ depending on the

food. Specifically in perishable food products like fish, their unique chemical composition

makes them prone to enzymatic degradation of proteins, lipids and trimethylamine-N-oxide

(TMAO) [2]. From the time the fish is caught, chemical processes begin to occur that could

eventually lead to food spoilage. The fish begins to degrade using its own enzymes and

then its proteins start to decompose. There is also significant bacterial deterioration that

- 4 -

NFC Food Spoilage Monitoring Device 2.2 pH

begins right after death. However, the effect of this bacterial growth is noticed only after

a particular point. As the microbial activities of the fish increase, the enzymes in the fish

cause it to soften [2].

In milk, a similar chemical process is experienced. Specifically oxidation, lipolysis,

proteolysis and syneresis occur during the food spoilage process. Generally, food spoilage

encourages the growth of microbes and yeast. The growth of these substances changes the

chemical composition of the food being tested making them more acidic or basic depending

on the food. Due to the change in the acidity of the food and the general occurrence of

microbial activities in food as spoilage occurs, pH could be used to monitor food spoilage [3].

2.2 pH

pH is a measure of the hydrogen ion concentration in a substance. It is the measure of

the acidity or alkalinity of a solution. Based on the definition by Bronsted and Lowry, an

acid is a substance capable of donating protons to other molecules or ions while a base is

a substance that is capable of accepting such protons [4]. All acid base reactions can be

written as follows:

HA+B BH+ +A (2.1)

Acid + base = conjugate acid of base B + conjugate base of acid HA+

pH can be measured in both aqueous and non-aqueous solutions using different scales.

pH measurements are very important in a multitude of applications and in our specific

case, it will be used to measure the level of food spoilage. Specifically, the pH of an aqueous

solution can be calculated using the following equation:

pH = log[H+] (2.2)

- 5 -

NFC Food Spoilage Monitoring Device 2.2 pH

Where [H+] is the concentration of hydrogen ions in mol/L.

In order to measure the pH of a solution, very precise pH meters or pH indicators could

be used. These pH meters give very specific results regarding the pH of the solution being

tested. They usually include a silver chloride reference electrode and a glass pH electrode.

Details about the electrodes will be covered later on in section 2.3. pH indicators on the

other hand give a visual determination of the pH of the material under test. pH indicators

undergo the following reaction:

HInd+H2O H3O+ + Ind (2.3)

Where HInd represents the acid form of the indicator and the Ind represents the

conjugate base of the indicator. The ratio of the conjugate base of the indicator to the acid

form of the indicator determines the colour of the solution and the eventual colour of the

solution determines its pH. There are various pH indicators that have varying transitional

pH levels. They can be used to detect varying levels of acidity or alkalinity of a substance.

The logarithmic pH scale used for measuring pH in aqueous solutions is shown below:

Fig. 2.1: The Logarithmic pH Scale

As shown in Figure 2.1, aqueous solutions with pH values below 7 are considered acidic

and those above 7 are considered basic. The higher the pH value of a substance, the more

basic it is and the lower the pH value of a substance, the more acidic it is. For example,

acidic items like apple juice have a pH of 2.9 while neutral liquids like distilled water have

- 6 -

NFC Food Spoilage Monitoring Device 2.3 Electrodes

a pH of 7.

2.3 Electrodes

Electrodes are solid conductors used in electrochemical measurements to obtain information

on the induced potential of an electrolyte [5]. They are used in cathode and anode pairs.

The silver/silver chloride electrode used in this design is a cathode and it acts as a reference

electrode that undergoes reduction in an electrolyte. The mixed-metal oxide electrode is

an anode and it acts as a sensing electrode, which undergoes oxidation. The following

expression is used to calculate the potential measured by electrodes:

E =RT

Fln(

aH+

(pH2/p0)1/2) =2.303RT

FpH RT

2Fln(

pH2p0

) (2.4)

Where E is the measured potential

R is the universal gas constant

T is the temperature (Kelvin)

F is the Faraday constant

aH+ is the activity of hydrogen ions

pH2 is the partial pressure of the hydrogen gas

p0 is the standard pressure.

A solid-state reference electrode and a mixed-metal oxide (MMO) sensing electrode are

used in this design. The mixed-metal oxide electrode is used for detection since the potential

across its terminal changes in response to fluctuations in pH level. The solid-state reference

electrode does not respond to changes in pH and acts as a reference for the MMO electrode.

A hydrophilic polymer called hydrogel is used to coat the electrodes. Hydrogel absorbs

gases that are released during food spoilage. It also acts as an electrolytic solution where

- 7 -

NFC Food Spoilage Monitoring Device 2.4 NFC Technololgy

ions are exchanged between the reference and mixed-metal electrodes to provide voltage

changes.

2.4 NFC Technololgy

Near Field Communication (NFC) is a wireless communication technology that is used to

exchange data in a short range. The NFC technology implemented in the food spoilage

monitoring device is based on inductive coupling. It comprises a transmitter coil and a

receiver coil. The alternating current in the transmitter coil generates a magnetic field

which couples to the receiver coil.

The amount of energy transferred through inductive coupling from the transmitter to

the receiver depends on the coupling factor between the two coils. The coupling factor

depends on the size of the coils and their relative positions, and is inversely proportional to

the cube of the distance between the coils [6] [7].

The quality of the coupling is also determined by the Q factor of the system. A higher Q

factor indicates that the amount of energy dissipated by resistive elements is small relative

to the amount stored by the reactive elements. The higher the Q factor, the more efficiently

energy is transferred to the receiver [6].

At greater distances less energy is coupled to the receiver, meaning that the energy

transfer must be more efficient in order for the same amount of energy to be stored. There-

fore the minimum operating distance between the coils depends on the Q factor. For our

purposes, the minimum distance between the transmitter coil and the receiver coil for guar-

anteed operation will be 3 cm.

2.4.1 NFC System

This food spoilage monitoring device comprises the following main components: the inter-

rogator and the NFC tag. The transmitter coil is part of the interrogator, and the receiver

- 8 -

NFC Food Spoilage Monitoring Device 2.4 NFC Technololgy

coil is one of the components of the tag. There are two approaches for establishing near

field communication that were considered.

One approach is to use an impedance analyzer which generates a sine wave that is

swept over a specified frequency range. The impedance analyzer accurately measures the

impedance of the circuit connected to its terminals. The interrogator coil is connected to

the impedance analyzer and the tag is placed nearby. The interrogator coil and the tag coil

couple thereby allowing the frequency response of the tag to be determined.

The other approach is to use time domain gating method, which is discussed in Section

2.5.

The resonant frequency of the NFC tag responds to voltage changes across the elec-

trodes. As seen in Figure 2.2, the NFC tag comprises a sensor coil, a varactor (C), and a

low pass filter.

Fig. 2.2: NFC Tag Diagram

The sensor and interrogator coils can establish Near Field Communication through

inductive coupling. The sensor coil can either be hand wound or milled on a printed circuit

board. When the sensor coil is in parallel with the voltage controlled capacitor (varactor),

a resonant circuit is formed and the resonant frequency can be determined using Equation

2.5.

- 9 -

NFC Food Spoilage Monitoring Device 2.5 Time Domain Gating Interrogation

f0 =1

2LC

(2.5)

Where f0 is the resonant frequency of the circuit,

L is the inductance of the coil and

C is the capacitance of the varactor.

A varactor is a voltage controlled capacitor that changes its capacitance in response to

changes in the bias voltage. In the NFC Tag, when the capacitance of the varactor changes,

the resonant frequency of the circuit changes as well. The interrogator coil interacts with

the sensor coil. The low pass filter consists of a resistor in series with a capacitor. The filter

prevents the electrodes from being affected by the output signal of the interrogator. The

cutoff frequency of low pass RC circuit is calculated based on the following equation.

fc =1

2RC(2.6)

Where fc is the cut-off frequency,

R represents the resistance and

C represents the capacitance.

2.5 Time Domain Gating Interrogation

Time Domain Gating (TDG) is a method of interrogation, used to obtain information from

sensors in wireless systems that employ technologies such as RFID or NFC. In systems

based on electrical resonance, TDG allows for the natural frequency of the sensor to be

determined. The main components of a TDG-based interrogator are a signal generator, a

switch, a switch controller, as well as an inductive coil in the case of NFC-based systems.

It also includes either an oscilloscope or lock-in amplifier, depending on the specific mode

of operation. The configuration of these components can be seen in Figure 2.3.

- 10 -


Fig. 2.3: Time domain gating-based interrogator block diagram

2.5.1 Operation

TDG-based interrogation has 3 phases: a transmit phase, a receive phase, and a short delay

phase in between transmitting and receiving. The timing of these phases is set by the switch

controller. Figure 2.4 shows a timing diagram of these 3 phases.

Fig. 2.4: Time domain gating timing diagram

During the transmit phase the signal generator is connected to the interrogator coil

via the switch. The energy sent from the interrogator coil couples to the sensor coil, and

the sensor coil begins to resonate. The interrogator then exits the transmit phase, and the

signal generator is disconnected from the interrogator coil. In the case of NFC systems,

the delay phase allows for energy stored in the coil, due self-resonance, to dissipate. For

RFID systems, the delay phase allows for any waves that are reflected by the environment

- 11 -


to pass by and not interfere with the interrogation. The interrogator then transitions to the

receive phase, and the energy stored in the resonant circuit of the sensor couples back to

the interrogator in the form of an exponentially decaying sine wave. There are two methods

of analyzing the received signal in order to obtain the resonant frequency: measuring the

amplitude and measuring the frequency.

2.5.2 Amplitude-Based Approach

The amplitude of the received signal depends on the frequency of the signal used to energize

the sensor. The closer the frequency of the wave generated by the interrogator is to the

resonant frequency of the sensor, the greater the amplitude of the received signal [7]. By

sweeping the interrogation frequency and determining which frequency results in the great-

est amplitude of the received signal, the resonant frequency of the sensor can be determined.

The amplitude is measured by feeding signal to a lock-in amplifier during the receive phase,

which outputs a DC voltage based on the amplitude.

2.5.3 Frequency-Based Approach

The frequency of the received signal is equal to the natural frequency of the sensor [7].

The interrogator operates at a fixed frequency, chosen to be close to the expected natural

frequency. The coil is connected to an oscilloscope during the receive phase, which allows

the frequency of the received signal to be measured. This frequency-based approach was

employed in this project.

- 12 -

NFC Food Spoilage Monitoring Device 3. Design Methodology

Chapter 3

Design Methodology

The following chapter details the design process of the NFC food monitoring system. In the

first section is a short description of how the various concepts introduced in Chapter 2 were

combined to create a wireless system for measuring pH. The subsequent sections discuss

the design and implementation of the different components of our project. The rationale

behind the decisions we made is provided, along with the results of the calculations and

simulations that were performed.

3.1 Introduction

In this project, the goal was to detect food spoilage through pH change. The gases that are

released by spoiling food will be absorbed by hydrogel, causing a shift in its pH. The voltage

generated across electrodes immersed in hydrogel will change, varying the bias voltage of

a varactor which results in a change in its capacitance. This change in capacitance alters

the natural frequency of the resonator formed by the varactor and the inductive coil of

the sensor. The natural frequency is determined by communicating with the sensor using

an interrogator employing NFC technology. Through interrogation, changes in the natural

frequency can be measured, allowing food spoilage to be detected.

- 13 -

NFC Food Spoilage Monitoring Device 3.2 Electrodes

3.2 Electrodes

3.2.1 Fabrication of Solid-State Reference Electrode

The solid state reference and the mixed-metal oxide electrodes were fabricated in the chem-

istry department at the University of Manitoba. The process began with the tip and sides

of silver wire being polished with fine grit sandpaper to ensure a clean silver surface. Then

+0.5 V was applied to the silver wire for 50 s in a 0.1 M potassium chloride (KCl) solution,

which formed a layer of silver chloride on the wire. Before coating the wire with some

immobilized electrolyte, its potential was measured in 3 M NaCl to check for accuracy and

determine if the wire is suitable for fabrication. An immobilized electrolyte was prepared by

saturating 12 mL of tetrahydrofuran (THF) with NaCl and then adding 0.4 g of polyvinyl

chloride (PVC). The silver-silver-chloride (Ag/AgCl) wire was dip-coated in the immobilized

electrolyte solution and then dried in a desiccator for 48 hours to evaporate the solvent.

After drying, the electrode was dip-coated with a protective polymer layer, three times,

to prevent leakage of chloride ions. The polymer used in this process was Nafion since it

provides increased reproducibility and stability compared to alternatives like polyurethane.

Next, the electrode was cured at 120 deg C for 1 hour, stored in a desiccator overnight and

then placed in 18 M water for 24 hours. The water was tested for chloride leakage, and

the electrode was stored in a desiccator [8].

3.2.2 Interfacing Electrodes to the Tag

Copper wires were attached to the electrodes, but conventional soldering presented a chal-

lenge, since heat generated by soldering irons (an average of 350 deg C) could damage the

temperature-sensitive electrodes. Conducting epoxy was used to attach wires to each elec-

trode since it provided the required stability and the process only required moderate baking

at about 65 deg C. Since the electrodes were designed to work in an acidic environment,

- 14 -

NFC Food Spoilage Monitoring Device 3.3 NFC Tag

exposed silver joints and conducting epoxy were prone to corrode. To prevent this from

happening, the exposed silver parts on the electrodes, and conducting epoxy were isolated

with an insulating epoxy coating. Finally a hydrophilic polymer called hydrogel was used

to coat the electrodes. The thickness of the coating was about 0.5 cm. Hydrogel was used

because it acts as an electrolytic solution where ions are exchanged between the reference

and mixed-metal electrodes to provide voltage changes. The hydrogel-coated electrodes can

be seen in Figure 3.1.

Fig. 3.1: Electrodes immersed in hydrogel

3.3 NFC Tag

3.3.1 Varactor

The varactor is one of the main components in the resonant circuit of the NFC tag. The

shift of the resonant frequency in the resonator depends on the change of capacitance

in the varactor. The electrodes in used in this design were expected to produce the a

bias voltage between 0 V to 0.4 V. Therefore the capacitance of the varactor needs to be

sensitive to changes in the bias voltage in this range. This will ensure that changes in the

electrode voltage cause distinguishable changes in the resonant frequency. The Philips BB2

low-voltage variable capacitance diode was selected because it satisfies the aforementioned

requirement as it provides 32.5 pF at 0.1 V bias voltage to 20 pF at 1 V bias voltage.

- 15 -


The graph of the varactor capacitance as a function of the bias voltage from Philips BB2

datasheet is shown in Figure 3.2.

Fig. 3.2: Varactor capacitance vs reverse bias voltage curve

3.3.2 Sensor Coil

Another main component of the NFC tag is the sensor coil. This coil establishes near field

communication with the interrogator, and behaves as an inductor which combined with the

varactor acts as an LC resonant circuit. However, there are no perfect inductors and there

is a turn-to-turn distributed capacitance called parasitic capacitance which exists in the

sensor coil. This parasitic capacitance makes the coil resonate at a self-resonant frequency

without being connected to the varactor. This self-resonant frequency is an important design

parameter because the desired sensing resonant frequency cannot be above the self-resonant

frequency. If the desired sensing resonant frequency is above the self-resonant frequency,

the sensor coil will act as a capacitor [9].

Figure 3.3 shows the inductance of a coil measured by the Agilent 4294A Precision

Impedance Analyzer. The inductive value rises sharply around 33 MHz and then abruptly

switches to a large negative value. Thus 33 MHz is the self-resonant frequency of the coil.

When the operating frequency is below self-resonant frequency, the coil acts as inductor

(positive L). When the operating frequency is beyond self-resonant frequency, the coil acts

- 16 -


Fig. 3.3: Inductance value of the sensor coil measured by the impedance analyzer

as capacitor (negative L). At the self-resonant frequency, the coil acts as a pure resistor.

We decided to build the coil with the self-resonant frequency about 30 MHz - 50 MHz

so the sensing resonant frequency of the resonant circuit in the design has to be much less

than this self-resonant frequency. We chose the desired sensing resonant frequency to be

13.56 MHz because it is in the ISM band. This frequency selection is also much lower than

the self-resonant frequency range of the sensor coil.

Based on the varactor selected and the specified sensing resonant frequency, the induc-

tance of the coil can be determined by Equation 3.1 below.

f =1

2LC

(3.1)

Where f is the sensing resonant frequency of the resonator (13.56 MHz),

C is average capacitance of the varactor (30 pF), and

L is the inductance of the sensor coil.

- 17 -


This results in a inductance value of 4.6 H, which is the important reference parameter

used to build the sensor coil.

Hand wound coil

In order to construct the sensor coil in the design, we decided to first build a hand wound

coil and then mill the final coil on the PCB afterwards. We decided to build the hand wound

coil as it was impractical to mill the initial testing PCB more than once, due to budget

limitations. The hand wound coil was relatively easy to construct and had no impact on our

budget. The purpose of building the first hand wound coil was to experimentally observe

the behaviour of the NFC tag and to confirm the implementation of our calculation above.

Finally, a printed spiral coil was implemented in the final design.

In order to build the hand wound coil however, we had to select a suitable core. The

most important considerations in core selection are summarized in Table 3.1.

Table 3.1: Core Selection

Min L Max L Type of Core Adjustable High Current Frequency Limit

20 nH 1 uH Air cored, self supporting Y Y 1 GHz

20 nH 100 uH Air cored, on former N Y 500 MHz

100 nH 1 mH slug tuned open winding Y N 500 MHz

10 uH 20 mH Ferrite ring N N 500 MHz

20 uH 0.3H RM Ferrite Core Y N 1 MHz

50 uH 1 H EC or ETD Ferrite Core N Y 500 MHz

1 H 50 H Iron N Y 10 KHz

These parameters were key design constraints throughout the core type selection. Due

to the specified sensing resonant frequency 13.56 MHz and the inductance of the sensor coil

4.6 H, the Table 3.1 above proves that the air core coil is best for our design.

Compared to a ferromagnetic core coil, the air core coil has a better Q-factor, greater

efficiency, greater power handling, and less distortion [10]. A copper wire was used to

- 18 -


construct the air core coil. The inductance value of the coil was determined by several

physical parameters such as the wire radius, the number of turns in the coil, cross-sectional

area of the coil and the influence of skin effect. The formula used to determine the initial

parameters in order to achieve the specified inductance value of 4.6 H is shown below.

L = 0.001N2r2/(228r + 254l) (3.2)

Where L is the inductance of the coil

N is the number of turns of the coil

l is the coil length in metres

r is the coil radius in metres

However, the coil, like all non-theoretical electrical components, is imperfect and due to

this imperfection, there are some typical errors in the formula. The best way to identify these

physical parameters is through trial and error after determining a starting point from the

formula. The desired dimensions were thus determined through repeated experimentation.

The details of these tests and the experimental results are found in Chapter 4.

Once the performance of the hand wound coil prototype met our design specifications,

we moved on to finding the physical parameters of the spiral coil that was eventually milled

on the PCB.

Printed Spiral Coil

The single layer planar spiral coil inductor was selected as the sensor coil that was eventually

milled on the PCB. The planar coil was chosen because it is mostly used in high frequency

applications and designed as traces on a circuit board [11]. The PCB used in this design

is copper clad and uncoated. Its dimensions are 114.4 mm by 76.2 mm and its thickness is

0.79 mm. The equation used to obtain the planar spiral coil is shown in Equation 3.3.

- 19 -


Lmw = K10n2davg

1 +K2(3.3)

where Lmw is the inductance of the planar spiral coil,

n is the number of turns,

is the fill ratio = (dout din) + (dout + din)

K1 and K2 are constants for a square layout and equal to 2.34 and 2.75 respectively

davg is the average diameter defined by davg = 0.5( dout + din).

This equation is called the modified Wheeler expression which is simple and provides

good accuracy. This expression, which has errors of 1 to 2%, represents a great improvement

over previously published expressions, which have typical errors of around 20% or more [12].

From the specification, we defined the dout = 5 cm and din = 1 cm. Solving Equation 3.3

with L = 4.6 H, we obtained the number of turns, 12. After the physical parameters of the

planar spiral coil were determined, Utiliboard software was used to create the schematic

layout of the PCB. The layout design was then exported to a Gerber file and sent for

manufacturing. Figure 3.4 shows the layout of the single layer planar square spiral coil.

Fig. 3.4: Single layer planar square spiral coil

- 20 -


3.3.3 Resonant Circuit

Fig. 3.5: An equivalent circuit diagram of the PH sensor and coupled interrogator coil

An equivalent circuit diagram of the pH sensor and coupled interrogator coil is shown

in Figure 3.5. The sensor coil is in parallel with a voltage controlled capacitor and the pH

sensor electrode. Ls is the series inductance of the sensor coil and Rs is the series resistance

of the sensor coil. C1 is the series capacitor capable of increasing the sensitivity to the

voltage changes. Cv is the voltage controlled capacitor (varactor) in the resonant circuit.

In the reverse bias state,

Cv = C0

(1 Vc

)(3.4)

Where C0 is the junction capacitance at zero bias,

is the junction built in potential,

is the doping dependent exponent,

Vc is the bias voltage across the varactor [7].

In order to minimize the effect of the resonant circuit on the varactor, we chose a

capacitor C1 = 1 nF which is much larger than Cv. Since this capacitor C1 is much

larger than the varactor capacitance, the equivalent input capacitance of the pH sensor

- 21 -

NFC Food Spoilage Monitoring Device 3.4 Interrogator

is approximately equal to the varactor capacitance thereby maximising the effect of the

varactor capacitance on the resonant frequency. R1 is the high impedance resistor which

ensures that most of the voltage change by the electrode is across the varactor. Therefore,

we chose R1= 10 G. R2 and C2 work together as a low pass filter which prevents the

signal generated by the interrogator from affecting the electrodes. We chose the cut-off

frequency to be 1.6 kHz which is much lower than our operational frequency. Thus, R2 =

1 M and C2 = 100 pF, according to Equation 2.6. R3 is the resistance developed by the

two electrodes, and V1 is the voltage across the two electrodes.

For

R3 (R2 +R1) (3.5)

Vc = V1 (3.6)

C1 C2 (3.7)

and a small interrogator source oscillation amplitude, the potential change across the elec-

trode can be monitored by obtaining the resonant frequency of the pH sensor circuit.

3.4 Interrogator

This section goes into details of the options considered in designing the interrogator seen in

Figure 2.3. We looked into designing a portable interrogator but eventually chose to design

a stationary interrogator. The design process is explained below.

3.4.1 Portable Interrogator

Bulky time domain gating-based interrogator setups already exist so we decided to start

our design by attempting to design a hand held device to perform the same function. In

order to do this, we needed to find smaller alternatives for the various components involved

- 22 -


in time domain gating that are described in Section 2.5. We began the design by breaking

down the interrogator section into two distinct parts: a transmitter and a receiver.

Transmitter

The transmitter was intended to replace the function generator in the current interrogator

setup. The goal of the interrogator was to produce a variable frequency sine wave specifically

in the range of 13 MHz to 14.5 MHz because our desired resonant frequency at the time was

13.56 MHz. For this purpose we considered a Voltage Controlled Crystal Oscillator (VCXO)

because its output frequency can be varied based on an applied control voltage. This would

allow us to adjust the frequency to ensure that it is close to resonant frequency of the sensor.

This would also enable us to begin constructing the interrogator immediately, rather than

waiting until the sensor had been fabricated and its exact resonant frequency determined.

In order to be able to implement the transmitter, the oscillator needs an accurate variable

voltage supply to obtain the required oscillation frequency.

Receiver

The next step in designing a portable interrogator was designing a device capable of mea-

suring the frequency of an incoming decaying sine wave received from the NFC tag. Several

approaches were considered in order to accomplish this. In the first method, the received

signal would initially be amplified, in order to increase its otherwise small amplitude. The

signal would then be passed through a high pass filter, which would cause the signal to

be attenuated by an amount related to its frequency. Next the signal would pass through

a diode to a capacitor, which would act as a peak detector and be charged to a voltage

proportional to the amplitude of the signal. This voltage would be recorded by an Analog

to Digital Converter (ADC). The recorded voltage reading could then be used to calculate

the frequency value of the received signal. A block diagram of this design can be seen in

- 23 -


the Figure 3.6.

Fig. 3.6: Portable Interrogator Receiver - design 1 block diagram

However, we realized that this method would be ineffective, as it requires the signal

received to always have the same initial amplitude. In reality the initial amplitude of the

received signal would depend on the distance and orientation of the sensor and interrogator

coils. It would be difficult to ensure that these two parameters were exactly the same for

every reading, making this method impractical.

The second method involved measuring the frequency directly. As in the first method,

the signal would first be amplified. It would then be fed to a comparator. The comparator

would convert the sine wave into a square wave signal by comparing the sine wave received

to ground. The square wave produced should have the same frequency as the sine wave

received if the signal received contains very little noise. The output of the comparator would

be connected to a digital counter Integrated Circuit (IC), which would record the number

of cycles the signal goes through. Knowing the number of cycles and the duration of the

interrogation allows the frequency to be calculated, which would be converted to a pH value

and displayed on an LCD. A block diagram of this method is shown in Figure 3.7.

- 24 -


Fig. 3.7: Portable Interrogator Receiver design 2 block diagram

We simulated this design using an exponentially decaying sine wave to model the in-

coming signal. Based on a discussion we had with Dr. Bridges, we set the initial amplitude

to be 100mV and the rate of decay to be 106 s1. However, due to the small amplitude

and rapid decay, the our circuit was only able to track the number of cycle for 3.6 us. The

precision with which we can calculate the frequency depends on the duration over which

we can count cycles, and we were concerned whether or not we would able to achieve the

precision necessary for our application.

Ultimately neither of these designs were implemented. After consulting with Daniel

Card and our supervisor, Professor Bridges, we determined that designing a portable inter-

rogator would take longer than the time we had and was outside the scope of our project

as outlined in our proposal. We moved on to implement a stationary interrogator.

3.4.2 Stationary Interrogator

The approach to the stationary interrogator involved using a time domain gating response.

It involves using a CW source to send out a single frequency as a pulse. The tag will pick up

the signal, couple to the resonator and start transmitting back. Since some energy is stored

in the resonator, the signal sent back would die down at the natural frequency. The rate at

which the frequency dies down depends on how much energy leaks out to the antenna and

how much energy goes to losses i.e. the Q factor of the resonator. Specifically, the Q factor

- 25 -


is related to the decay rate using the equation below:

Q =n

2 (3.8)

Where is the decay rate in s1 and

n is the natural frequency

Therefore, a reasonably large Q factor is needed for the NFC tag and a much smaller

frequency is required for the interrogator coil.The time domain gating method could be

achieved by energizing the NFC tag first, turning off the source and then looking at the

received response on an oscilloscope continuously. Once this has been repeated a few times,

the resonant frequency of the tag could be obtained using the average of the received signals.

By doing this, the noise from the surroundings would be considerably reduced because the

resonant circuit of the NFC tag slowly releases its energy as it returns to the interrogator

coil. Since the surroundings do not store energy like a resonant circuit does, noise would

be reduced drastically.

Once we decided to implement the time domain gating method, we began our design by

analyzing the built-in interrogator in the RF Lab. We wanted to understand how exactly

the setup worked and how to go about designing a similar device to perform the same

function. The interrogator setup is shown in Figure 3.8. It includes switches, a switch

controller, an inductive coil, a function generator and an oscilloscope.

- 26 -


Fig. 3.8: Stationary interrogator setup block diagram

The ZYSWA-2-50DR switches were used in this project because they were readily avail-

able and met our requirements. These switches require a +/-5 V DC supply and have a

wide bandwidth of 5 GHz which is much higher than our operating frequency range of 12.37

MHz - 13.48 MHz. They also have a low insertion loss of 1.1 dB at our operating frequency

range.

The implementation of a stationary interrogator required the design of a switch con-

troller, the use of a voltage regulator and some code for data acquisition. All the other

necessary components were provided by the university, as described in our proposal.

- 27 -


Switch Controller

The switch controller was built using 555 timers, a delay chip and a crystal oscillator. In

order to implement the switch controller we considered using a variety of techniques but

we eventually chose to use 555 timers to implement the switch. We chose to use 555 timers

because we could easily change the duty cycle of our waveforms and also because they are

cheap and available. Our goal was to achieve a 50% duty cycle waveform and a 20% duty

cycle waveform with a frequency of 25 kHz and a delay of about 1us between them. We

wanted to achieve these specific waveforms because these timings had been used in the

existing TDG interrogator setup in the RF lab which had been confirmed to work in our

application (see Section 4.2.5).

However, one issue is that, in their typical configurations, these chips cannot generate

signals with a duty cycle less than 50%. We overcame this by instead generating an 80%

duty cycle signal and passing it through an inverter. We began by considering the operation

of the timers in astable configuration, where the oscillation is controlled by the values of 2

external resistors and a capacitor: RA, RB and C. The frequency of oscillation is determined

according to Equation 3.9 .

f =1.44

(RA + 2RB)C(3.9)

The duty cycle is determined according to Equation 3.10.

D = 0.693f [(RA +RB)C] (3.10)

With these two equations, we were able to calculate the values of RA, RB and C required

for the desired operation for both of the timers. For the 80% duty cycle timer, the values

of RA, RB, and C were 3.47 k, 1.15 k, and 10 nF, respectively. For the 50% duty cycle

timer the values were calculated to be 100 , 2.45 k, and 11.5 nF, respectively. A circuit

- 28 -


containing two 555 timers configured with the components listed above was simulated using

Multisim. The output of this circuit can be seen in Figure 3.9.

Fig. 3.9: Simulation of 555 timers in astable configuration

The frequency and duty cycle of the two waveforms met our requirements. However, the

phase relation between the two signals was constantly changing, due to a slight frequency

mismatch. This mismatch was a result of the component values not being precise enough

to ensure an exact frequency for both timers. This effect would be even more pronounced

in a real-world implementation, as there would be a degree of error in the values of the

components used.

To overcome this problem, a second design was considered. The timers would instead

be operated in the monostable mode. In this configuration the duty cycle is determined

the values of a resistor and a capacitor, RA and C, while the frequency is controlled by an

external clock signal. The duty cycle is determined by Equation [3.3]

D = 1.1RACf (3.11)

Where f is the clock signal frequency.

For the 80% signal RA and C were calculated to be 2.91 k and 10 nF, respectively.

The values calculated for the 50% duty cycle signal were 1.82 k and 10 nF, respectively.For

the role of the external clock signal, a crystal oscillator was chosen, due to their accuracy

- 29 -


and stability. However, a 25 kHz crystal oscillator was not available, so we instead chose

a higher frequency that would be divided using D flip-flops. Each flip-flop in series divides

the frequency in half, so with 8 flip-flops and a 6.144 MHz oscillator, a 24 kHz signal was

obtained, which was sufficiently close to our target frequency of 25 kHz.

This new design was simulated in Multisim, and the output can be seen in Figure 3.10.

Fig. 3.10: Simulation of 555 timers in monostable configuration

This configuration resulted in the production of synchronized 24 kHz frequency 50%

and 20% duty cycles, thereby eliminating the problem of frequency mismatch associated

with the previous design. Therefore we decided to use monostable configuration described

above in our design of the switch controller.

Now that we had obtained the required waveforms, we looked into the delay portion of

the switch controller. To obtain the delay between the two waveforms, we first considered

using 555 timers. However, the delays produced by 555 timers start from the order of 10

ms and the delay we wanted to obtain was in the order of 1 s. Next, we looked into

programmable delay chips and we found the Linear Technology chip LTC6994-1 chip that

enabled us to program the exact delay we wanted using excel. The program defined the

delay as the amount of time you wanted the rising edge of the waveform to be shifted by

- 30 -


without affecting the falling edge.

Once we realized how the delay chip functioned, we revisited the design of the syn-

chronous waveform design of the 50% and 80% designs and decided to use the square wave

output of the 8 flipflops as our 50% duty cycle. We also chose to generate a rectangular

waveform that is high for 30.27 s instead of 80%(33.33 s). This dropped our duty cycle

to 73.728%. We chose to reduce the duty cycle so that we could move the rising edge of

the new 73.728% duty cycle waveform to be about 1 s after the square waveform therefore

we chose to delay the rising edge by 21 s. Using a simulation program called LTspice

IV, We were able to simulate the function of the chip and verify that the results of the

chip matched the required specification. The diagram of the delay chip with the resistor

and capacitor values selected are shown below in Figure 3.11. This diagram also shows the

resulting waveforms.

Fig. 3.11: Configuration of the Delay Chip

By combining the delay chip, crystal oscillator and 555 timers, we built the switch

controller on two prototype boards. The built switch controller is shown in Appendix A.

- 31 -


Fig. 3.12: Simulation Results of the Delay Chip

Sensor Resonator Substitute

In order to successfully design and test the interrogator, we would need to observe how

it functioned in conjunction with a resonator. However, due to the fact that work on the

interrogator was being performed concurrently with the work on the sensor, we could not

rely on the sensor resonator being completed beforehand. To overcome this, we designed a

resonator that would act as a stand-in for the final sensor resonator. An inductive coil was

made by winding wire into a 5 cm diameter, 5 turn circular loop. The inductance of this

coil as measured by the impedance analyzer was 3.14 H. This coil was soldered to a 43 pF

capacitor in order to achieve a nominal resonant frequency of 13.7 MHz, in order to obtain

a resonator that would be suitable to use as a substitute for the resonant circuit of the pH

sensor. The parameters of the substitute resonator were later verified experimentally, as

seen in Section 4.2.5.

Data Acquisition System

The first step in developing a system to record and analyze the data received from the

interrogator was to find a method of capturing the signal displayed on the oscilloscope and

save it to a computer. To do this we utilized a piece of software called Openchoice, which

- 32 -


is made available online by the manufacturers of the oscilloscope, Tektronix. Openchoice

enables a computer to interface with the oscilloscope via USB, allowing the user to capture

and save a waveform as a table of data points, plotting voltage vs. time. The saved file also

includes the frequency at which the signal was sampled.

Next, a method for analyzing the obtained data was required. We chose to use a

program called MATLAB for this purpose, as the group had experience using it for signal

processing applications in prior courses. We then needed to write code in MATLAB that

would import the data saved from the oscilloscope, determine the frequency of the signal,

and then output the value.

The first method for determining the frequency was to count the number of times the

signal crossed the x-axis, which is equivalent to the number of times the signal voltage is

equal to zero. Because a sine wave is equal to zero twice per cycle, the number of cycles can

be calculated by dividing the number of zero-crossings by two. The duration over which

the signal was sampled can be calculated by dividing the number of data points by the

sampling frequency. The frequency of the signal can then be calculated based on Equation

3.12 .

f = N/Ts (3.12)

Where f is the frequency,

N is the number of cycles, and

Ts is the sampling duration.

When attempting to count the zero-crossings, we discovered it does not work to simply

count the number of data points at which the voltage is equal to zero. This is because

it is unlikely that a sample will occur at precisely the moment that the signal is equal to

zero, and instead will most likely sample the signal when it is just above or below zero.

Instead, the number of zero-crossings was determined by counting the number of time the

- 33 -


voltage changed from positive to negative, or vice versa. However, we realized that the

zero-crossing method would not be suitable for our application. The reason for this is that

in real-world situations there will always be some noise in the signal. Noise causes spurious

zero-crossings, which in turn results in an incorrect measure of frequency.

The next attempt involved a Fourier transform, which converts a time-domain signal

into the frequency domain, resulting in a function of amplitude versus frequency. The

frequency is then determined by finding the maximum of this function and outputting the

corresponding frequency value. This approach grants immunity to noise, so long as its

amplitude is less than the amplitude of the received signal. It also allows for the possibility

of implementing simple digital filtering by removing parts of the frequency-domain function

that lie outside of our frequency range of interest. This kind of filtering would allow the

data acquisition system to function even in the presence of noise with an amplitude greater

than that of the signal, provided that the noise was outside of the frequency range we are

working in.

The operation of the Fourier transform-based method was simulated by using a computer-

generated sine wave in the place of signal received by the interrogator. This generated sine

wave had an arbitrary frequency of 10 Hz and amplitude of 1 unit. Random noise with an

amplitude equivalent to 25% of the sine wave amplitude was then added. A graph of the

input sine wave can be seen in Figure 3.13, along with the same signal converted to the

frequency domain in Figure 3.14. The program correctly identified the frequency value as

being 10 Hz.

- 34 -


Fig. 3.13: A: 10 Hz sine wave with noise in time domain

Fig. 3.14: B: 10 Hz sine wave with noise in frequency domain

Power Supply

In order to power the switches and the switch controller circuits used in this design, we

needed to provide precise +5 V , -5 V and ground supplies. To obtain these values, we

decided to use the variable +-15 V supplies on a bread board connected to voltage regulators.

We chose to use the LM7805C +5 V voltage regulator and the MC7905C -5 V voltage

regulator because they were available at the University. The schematics used in building

- 35 -


the voltage regulators are shown in Figures 3.15 and 3.16.

Fig. 3.15: Schematic of the +5 V voltage regulator

Fig. 3.16: Schematic of the -5 V voltage regulator

- 36 -

NFC Food Spoilage Monitoring Device 4. Testing Procedure and Results

Chapter 4

Testing Procedure and Results

In this chapter we explain the testing that was performed in order to verify the operation

of our design. We begin by providing a list of the different pieces of equipment that were

used, along with a brief description of each. We then move on to the testing performed on

the various components that comprise our system. We conclude with the tests performed

on the system as a whole, with all the sub-components having been integrated. Within each

test entry we start by outlining the reason for why the test was performed, as well as the

necessary preparation. We then proceed to describe the steps taken in the test procedure,

followed by the results that were obtained. Finally, each entry ends with a discussion of

what conclusions were drawn based on the outcome of the experiment.

4.1 Test Equipment

4.1.1 DC Generator

A DC generator was used to emulate the biased voltage produced by the set of electrodes.

The power supply model that was used was the Hewlett-Packard 6115A Precision Power

Supply.

- 37 -

NFC Food Spoilage Monitoring Device 4.1 Test Equipment

4.1.2 Impedance Analyzer

An impedance analyzer is capable of obtaining the resonant frequency of the resonant

circuit by finding the frequency at which impedance is maximum, as well as measuring the

inductance and resistance of the coil. The impedance analyzer model that we used was the

Agilent 4294A Precision Impedance Analyzer, which had an operational frequency range of

40 Hz to 110 MHz. This range covered our desired operating frequency. The impedance

analyzer is also capable of generating an equivalent circuit model to fit the data obtained.

4.1.3 Function Generator

A Function Generator was used to send small voltage sinusoidal signals to the interrogator

coil. The model of the function generator is Stanford Research Systems Model DS345. The

DS345 generates sine waves up to 30.2 MHz which covers our desired operating frequency.

4.1.4 Oscilloscope

An oscilloscope was used to display the signal received by the interrogator coil, and the

screen capture obtained was further analyzed. The model of the oscilloscope used is the

Tektronix TDS 2024B. The oscilloscope has a USB host and device ports that enable re-

movable data storage and PC connectivity.

4.1.5 Pulse Generator

A pulse generator was used to generate the TTL rectangular waves for the switches during

initial testing. The model of pulse generator is Quantum Composers Model 9618. The pulse

generator has 8 independent output and multi-channel capability.

- 38 -

NFC Food Spoilage Monitoring Device 4.2 Individual Component Testing

4.1.6 Digital Multimeter

An Agilent 3401A digital multimeter was used to measure voltage changes between the

mixed metal oxide and reference electrodes. Most common multimeters have an input

resistance of 10 M and the electrodes used in this design have a resistance of about 2

M. Since the both resistances are comparable, a phenomenon called loading effect occurs,

which causes inaccurate voltage readings when measurements are taken. The Agilent 3401A

multimeter was chosen because its input resistance can be changed to a relatively high 10

G.

4.2 Individual Component Testing

In this section we discuss the testing that was performed on the individual components that

together make up our final system design. These tests were undertaken in order to confirm

the operation of the various elements, allowing us proceed with combining the components

into the complete food spoilage monitoring device.

4.2.1 Electrodes testing

Purpose

The purpose of this test was to determine the response of the electrodes to gases of differing

pH levels. The voltages across the electrodes were recorded for an acidic gas, a basic gas

and neutral pH environment.

Setup

The setup for the testing process is shown in Figure 4.1. The electrodes were enclosed in a

glass chamber containing an aqueous solution. The MMO electrode was connected to the

high impedance input of an Agilent 3401A digital multimeter and the Ag/AgCl reference

- 39 -


Fig. 4.1: Complete test setup with multimeter attached (voltage shown is still changing)

electrode was connected to the ground input.

Procedure

De-ionized Water and Acetic Acid

The hydrogel-coated electrodes were enclosed in a glass chamber containing deionized

water and left for three days. This was done to allow the voltage between the electrodes

stabilize before any further testing. White vinegar contains 5% acetic acid, (as indicated on

the commercial vinegar used in this test) so it was used as the pH solution in this experiment.

After the voltage between the electrodes stabilized, deionized water was quickly replaced

with vinegar. Replacing deionized water with vinegar was done quickly to prevent acetic

acid from escaping. The voltage between the electrodes was monitored with a multimeter.

Ammonia

After the voltage between the electrodes settled at a value corresponding to the pH of

acetic acid, measurements were taken and the setup was taken down. The hydrogel coating

was removed and the electrodes were washed with de-ionized water, and left to dry. The

electrodes were recoated with hydrogel and isolated in the glass chamber containing deion-

ized water. After three days of stabilization, deionized water was quickly replaced with

- 40 -


Ammonium Hydroxide (NH4OH). Ammonium Hydroxide was used in this experiment be-

cause it contains 1% ammonia. The voltage changes between the electrodes were monitored

afterwards.

Results

The following results were obtained after testing:

Table 4.1: Voltage responses of the electrodes to varied pH levels

Substance pH Level (in aqueous solution) Voltage (mV)

Acetic acid 2.4 240

De-ionized water 7 104

Ammonia 11.6 -29.4

Using the data shown in Table 4.1, the following plot was generated:

Fig. 4.2: pH level vs. voltage

From Figure 4.2 above, we obtained V0, the voltage at 0 pH value, to be 309.84 mV.

The slope of the regression line was found to be -29.283 mV/pH, which gives equation 4.1

shown below:

V (mV ) = 29.283(pH) + 309.84 (4.1)

- 41 -


The operating pH range of the electrodes is approximately 2.4 to 11.6 and with this

information we determined the operating voltage range to be -29.84 to 239.56 mV.

Conclusion

This experiment examined the response of the electrodes to acidic, neutral and basic gases.

The electrodes showed a linear voltage response to varied pH levels, and this conformed

to our expectations. The operating voltage range we determined is wide enough to cause

a corresponding change in the tags resonant frequency that can be detected by the inter-

rogator.

4.2.2 Sensor Coil testing

Hand Wound Coil Testing

Purpose

The purpose of this test was to determine the physical parameters and dimension of the

hand wound air core coil in order to achieve the specified inductance value of 4.6 H and a

high Q factor. This coil was used to build the NFC tag prototype.

Setup

The inductive coil was plugged into the test fixture which was attached to the Agilent 4294A

Precision Impedance Analyzer. Figure 4.3 illustrates the basic setup for the testing of the

hand wound air core coil.

- 42 -


Fig. 4.3: Setup of the testing of the hand wound air cored coil

Procedure

The impedance analyzer is designed to measured the complex impedance Z and or R+jX

instead of measuring inductance L and capacitance C directly. Where Z is the complex

impedance, is the associated phase angle, R is the real part of Z, and X is the imaginary

part of Z. If the impedance analyzer is set to display L, the value of L will be calculated by

equation 4.2.

L = X/ (4.2)

where is the angular frequency in radians/sec. X can be positive or negative. When

X is positive, X is considered as inductive reactance. When X is negative, X is considered as

capacitive reactance. In other words, from equation 4.2, capacitance is negative inductance.

When the operating frequency is below the self-resonant frequency, the inductive coil

appears to be inductive. When the operating frequency is above the self-resonant frequency,

the inductive coil appears to be capacitive. At the self-resonant frequency, the impedance

- 43 -


reaches a maximum and is purely resistive. If the impedance analyzer is set to display

R, the position of the peak value of the resistance (the real part of Z) corresponds to the

self-resonant frequency of the inductive coil. This is the main principle used to obtain the

resonant frequency of the inductive coil from the impedance analyzer.

In the first setup, we wound the copper wires with 11 turns and a 4 cm diameter.

Frequency sweeps were done from 10 MHz to 20 MHz. The impedance analyzer was set to

display Ls-Q model. We checked the value of L that corresponded to the specified sensing

resonant frequency 13.56 MHz. This test was then repeated by adjusting the number of

turns and the cross sectional area of the air core coil until the value of L was closed to 4.6

H at the specified sensing resonant frequency 13.56 MHz. The Q factor of the air core coil

was also measured with the impedance analyzer.

Results

After testing the hand wound air core coil, we found that increasing the number of turns

and the cross sectional area of the air core coil increases the inductive value. In order to

meet our design specification of 4.6 H, we repeatedly adjusted these two parameters. The

Q factor and the final dimensions of the air core coils are shown in Table 4.2.

Table 4.2: Dimensions of the Air Core Coils

Hand Wound Air Core Coil

Q 122

Diameter of the coil 3.6cm

Number of turns 7

Figure 4.4 shows the built hand wound air core coil. In Figure 4.5, we see the inductive

value of the coil is 4.33 H at the specified sensing resonant frequency 13.56 MHz. This

inductive value is close to our specified value 4.6 H so the design of the air core coil was

- 44 -


Fig. 4.4: Hand wound air core coil

acceptable.

- 45 -


Fig. 4.5: Inductive value of the hand wound coil

Figure 4.6 shows the self-resonant frequency of the coil. As expected the sensing res-

onant frequency 13.56 MHz of the resonant circuit is much less than this self-resonant

frequency.

Fig. 4.6: Self-resonant frequency of the hand wound coil

Printed Spiral Coil Testing

Purpose

The purpose of this test was to confirm that our milled PCB coil met specifications. The

inductance value of the coil should be 4.6 H at 13.56 MHz (operating frequency) and the

Q factor of the coil should be high.

- 46 -


Setup

Two wires were attached to both ends of the printed spiral coil, and the other ends of these

wires were plugged into the test fixture of the Agilent 4294A Precision Impedance Analyzer.

Figure 4.7 illustrates the basic setup for the testing of the printed spiral coil.

Fig. 4.7: Milled PCB coil

Procedure

The testing procedures for printed spiral coil are similar to the testing procedures for the

hand wound air core coil. The impedance analyzer was set to display Ls-Q model. The fre-

quency sweeps were from 10 MHz to 20 MHz. We checked the value of L that corresponded

to the specified sensing resonant frequency 13.56 MHz. The Q factor of the printed spiral

coil was also measured with the impedance analyzer.

- 47 -


Results

After testing the printed spiral coil, we found that its inductive value was 4.6 H at the

specified sensing resonant frequency 13.56 MHz and its Q factor was 120. The results

of testing satisfied our specifications, and verified the accuracy of the modified Wheeler

expression that was used to calculated the physical parameter of the single layer planar

spiral coil.

Conclusion

The quality of the inductive coil was determined through the measurements from impedance

analyzer. Both the hand wound air core coil and the printed spiral coil had a high Q factor

and their inductance values were close to the specified inductance value of 4.6 H at the

operating frequency of 13.56 MHz. Thus, the designs of the inductive coils were acceptable.

4.2.3 NFC Tag Testing

Tag with a Hand Wound Air Core Coil

Purpose

The purpose of this test was to experimentally observe the behaviour of NFC tag. This tag

was used as a prototype to confirm the implementation of our design.

Setup

An interrogator coil was connected to the impedance analyzer and the NFC tag was 4.4

cm away from the interrogator coil on the impedance analyzer. The DC generator was

connected across the varactor. Figures 4.8 and 4.9 illustrate the setup for testing the NFC

tag with the hand wound coil.

- 48 -


Fig. 4.8: Top view NFC tag testing set up

Fig. 4.9: Front view NFC tag testing set up

- 49 -


Procedure

Using the DC generator, the bias voltage across the varactor was varied from 0 V to 0.4

V with 0.1 V divisions. The resonant frequencies of the circuit were obtained using the

impedance analyzer.

Result

The sensing resonant frequency shifts from 13.15 MHz to 14.45 MHz. This resonant fre-

quency shift is due to the 0.1 V to 0.4 V change in the bias voltage across the varactor. The

calculated Q factor of the resonant circuit is 134. The largest operational distance between

the sensor coil and interrogator coil is 6.1 cm. The result of testing met our expectation.

These results provided firsthand data of NFC tag performance. The data was also used as

a reference for building the tag with printed spiral coil.

Conclusion

The NFC tag with hand wound coil satisfied our design specification. However, from the

aforementioned test, we realised that the hand wound coil is very sensitive to movement.

This problem was overcome by using the milled PCB coil for our final design.

Tag with printed spiral coil

Purpose

The purpose of this testing was to determine the relationship between the bias voltage

across the varactor and the resonant frequency of the resonant circuit. It was also aimed

at measuring the Q factor of the resonant circuit and determining the operational distance

between the sensor coil and interrogator coil. This tag was implemented in the final design.

- 50 -


Setup

An interrogator coil was connected to the impedance analyzer and the NFC tag wass 4.4

cm away from the hand wound coil on the impedance analyzer. The DC generator was

connected across the varactor. Figures 4.10 and 4.11 illustrate the ove

designing a food spoilage monitoring device that uses near field

Documents