CHAPTER 1

INTRODUCTION

The essence of engineering lies in making improvements in every walk of life, thus

taking the world from ‘what is’ to ‘what could be’. The feats in science and technology

have, over the years, made it possible to better understand the world we live in and

build insights as to how it can be elevated through engineering. To make even the

smallest improvement to any process, product or system, in any field, the most efficient

approach is to quantify the parameters of interest. As the eminent engineer and natural

philosopher Lord Kelvin put it, ‘If you cannot measure it, you cannot improve it’.

Practical experience sets certain criteria that the chosen measurement method needs to

fulfil because a measurement is useful only if the adopted method gives a numerical

value that lies within the acceptable percentage of error for the intended application.

This understanding has given birth to the ever-expanding role of sensors in every field

of human endeavor.

Sensors are to machines what the sense organs are to human beings. They translate

external physical stimuli into a form that is understandable to the machine. Thus,

practically speaking, a sensor is a device that converts a physical quantity to an

electrical quantity [1]. There are a wide variety of sensors to measure every parameter

of significance. They can be classified based on different criteria: (a) Power supply –

active or passive, (b) Output signal – analog or digital, and (c) Operating mode –

deflection or null mode [2].

The classical sensor unit, as shown in Fig. 1.1, consists of the sensing element, followed

by a signal conditioning block, and an analog-to-digital conversion (ADC) block. The

measurand manipulates the sensing element. The resultant signal embeds information
 Fig. 1.1 Classical sensor unit

about the measurand. This is extracted by the signal conditioning block that typically

provides an analog output. It is subsequently digitized by the ADC block, whose digital

output is generally read by a microcontroller. It is fair to say that a sensor is only as

good as its signal conditioning circuit (including the ADC), as it is this block that

mainly determines the overall sensitivity, linearity, power consumption, accuracy, and

ease of interfacing the sensor with digital systems for automation, analysis, and control.

A sensor unit that fulfils its desired objectives to the highest degree is critical, especially

when it is present in the feedback path of process control systems, autonomous systems,

and systems with self-diagnostic capability [3]. Even the smallest error in the

performance of this feedback element can undermine the entire system.

In several applications, miniaturized sensors are preferred. Thanks to the advances in

the microelectronic design, the blocks of the sensor unit shown in Fig. 1.1 can be

integrated into a single chip. The sensing element can be designed and fabricated using

Micro-Electro-Mechanical Systems (MEMS) technology. Similarly, self-powered

sensors utilizing energy harvesting technologies (e.g. a photovoltaic panel) are needed

to deploy sensors in locations where access to electrical power is not available or

limited. These approaches help to realize smart and/or autonomous sensors that are

important for the deployment wireless sensor networks and Internet of Things (IoT).

In the current era of robotics, machine learning, and artificial intelligence, the field of

sensor development is driven by the ever-growing demand for low-cost, easy to install,
2
robust sensors that offer high performance. This need is answered by the non-intrusive,

non-contact sensors. Especially while dealing with harsh environments, or in cases

where the measurand is physically inaccessible, or when contact with the measurand

degrades the measurement, the only viable solution is the non-contact approach [1].

The absence of the critical requirement of direct contact, between the measurand and

the sensing element, adds to the reliability and versatility of this sensing method.

Additionally, a non-contact sensing mechanism, being less prone to wear and tear, can

ensure the long life of the sensor, and impact the overall cost of the measurement system

positively [4].

Among the non-contact sensing paradigms available to the design engineer, one of the

most popular is the capacitive sensing approach. Capacitive sensors are known for their

simplicity in construction, robustness, high sensitivity, low-power consumption, and

accuracy [4]. The electrodes are available in many configurations, such as single-

element (a pair of electrodes, or a single electrode and ground), differential (three

electrodes), or with multiple-electrodes (more than three). These can be in the parallel-

plate form or the cylindrical or hemispherical shapes. They are used in consumer

electronics as touchscreens on smartphones, and other human-machine

interfaces (HMI) including touch-buttons on certain washing machines and

refrigerators [5]. Capacitive sensors are used in automobiles for obstacle-detection

while reversing the vehicle, to facilitate key-less entry [6], and in the HMI stereo and

media center [7]. In the automobile industry, the role of MEMS-based capacitive

sensing is also very significant, for instance, to measure the tire pressure and the

acceleration caused by a collision. In industrial applications, this sensing technology is

used for liquid level detection [8], differential pressure sensing [9], and monitoring

harsh environments [10]. These sensors are used extensively in healthcare [11], for non-
3
contact blood pressure measurement, sweat analysis, respiratory measurement, and in

agriculture for soil-moisture detection [12]. They are used in scientific research for

position [13], pressure [14] and humidity [15] measurements, and material analysis

[16]. Further, the application of capacitive sensors extend to electrostatic

accelerometers in space research [17], and robotics [18].

1.1 CAPACITIVE SENSOR

In capacitive sensors, the change in the measurand is translated to a variation in

capacitance, which the sensor senses and processes to give an equivalent electrical

output. The output can be in the form of a voltage, current, or a quasi-digital signal [19],

which is subsequently converted to a digital value in the final stage. The capacitive

sensitive element consists of two or more conducting electrodes, separated by an

insulating dielectric. The capacitance of the sensor is dependent upon the physical

dimensions of this element, number of electrodes and the material properties of the

dielectric material present between the electrodes [4]. A brief overview of some of the

widely used electrode configurations is given below.

1.1.1 Single-Element Parallel-plate Capacitive Sensor

This sensor typically consists of two conductor plates, placed parallel with respect to

each other. The capacitance 𝐶, which is the ability of a capacitor to store electrical

charge, is calculated by (1.1).

𝜖0 𝜖𝑟 𝐴
𝐶= (1.1)
𝑑

(1.1) is characterized by the overlapping area of the two plates, 𝐴, the dielectric constant

of the material between the plates, 𝜖𝑟 , the permittivity of free space (8.85 × 10-12 F/m),

𝜖0 , and the separation between the plates, 𝑑. Though this equation ignores the fringing

4
effect, it is an acceptable approximation as long as the distance (𝑑) between the plates

is small compared to the other dimensions of the plates such that the field in the

capacitor over most of its area is uniform.

The measurand is sensed based on its manipulation of 𝐴, 𝜖𝑟 , or 𝑑 [4]. A single-element

capacitive sensor that relies on the variation of capacitance due to change in either 𝐴 or

𝜖𝑟 exhibits linear sensor characteristic, as can be garnered from (1.1), while that which

is based on change in 𝑑 has inverse characteristic. These different sensing modalities

make the capacitive sensor a versatile one. Some interesting applications of the single-

element capacitive sensor include proximity sensing [20] or flow [21] measurement,

touchscreens [22] and respiratory rate measurement [11]. Some of these sensors use

planar electrode configuration which is relatively easy to manufacture.

1.1.2 Differential Capacitive Sensor

The simplified structure of a differential capacitive sensor (DCS) used in a differential

pressure sensing application is given in Fig. 1.2. Here, PR1 is the pressure at the input

port on the left hand side and PR2 is the pressure at the input port on the right. These

pressure ports shown in the diagram are meant for non-conducting fluids. 1, 2, and 3

are conducting plates. While 1 and 2 remain fixed, 3 is movable. The plates are

electrically insulated from each other. P, Q, and R are the terminals connected to the

plates 1, 3, and 2, respectively. The electrical equivalent of the DCS is given in Fig. 1.3

with the two sensing capacitances, 𝐶1 and 𝐶2 , sharing a common electrode. 𝐶1 and 𝐶2

vary with the parameter being sensed.

In the case of the differential pressure sensor in Fig. 1.2, when pressures PR1 = PR2, the

distance between plates 1 and 3 is the same as that between 2 and 3. So, 𝐶1 = 𝐶2 = 𝐶0 .

This indicates that the measurand, the differential pressure (PR1 - PR2) is zero. In case
5
Fig. 1. 2 The simplified structure of a differential capacitive sensor (DCS) used
to sense differential pressure. ‘i’ indicates the insulation layer.

Fig. 1. 3 Equivalent circuit of differential capacitive sensor

PR1 > PR2, then the differential pressure (PR1 - PR2) will be non- zero, and the middle

plate will move. Then the distance between the plates 1 and 3 will be greater than that

between 2 and 3, resulting in 𝐶1 < 𝐶2 . The direction of change in capacitance in 𝐶1 due

to the measurand is opposite to that in 𝐶2 , causing these sensors to be popularly known

as push-pull type capacitive sensors.

In a DCS, the change in capacitance with respect to the measurand, x, can be linear or

non-linear (inverse) as given in (1.2) and (1.3), respectively, where 𝑘 is the sensor

constant, and 𝐶0 is the nominal capacitance of the sensor [23]. The sensor shown in

6
Fig. 1.2 has an inverse characteristic as the distance between the plates changes with

respect to the measurand. Similar to the single-element sensor discussed in sec. 1.1.1,

for the DCS also the characteristic is linear, as in (1.2), in case of parallel-plate

arrangement of the electrodes which utilize the change in the area of overlap between

the plates, or change in the dielectric constant, with respect to the measurand.

𝐶1 = 𝐶0 (1 ± 𝑘𝑥) and 𝐶2 = 𝐶0 (1 ∓ 𝑘𝑥) (1.2)

𝐶
0 0 𝐶
𝐶1 = (1∓𝑘𝑥) and 𝐶2 = (1±𝑘𝑥) (1.3)

1.1.3 Applications of Capacitive Sensors

Capacitive sensors pervade varied fields such as consumer electronics, industry,

automobiles, agriculture, medicine, and research [5]-[18]. The key advantages of

capacitive technology are its simplicity, ease of fabrication, low power consumption,

high sensitivity, accuracy and resolution over the measurement range [4]. They are

employed in a multitude of applications ranging from touchscreens, proximity and

position sensing, food-quality assessment, occupancy detection, water-quality

monitoring, and in the measurement of pressure, flow, liquid-level, and humidity. A

few of these applications are briefly discussed in the following sub-sections.

1.1.3.1 Touchscreens

The prevalent touch technologies can be categorized into six basic types – capacitive,

resistive, acoustic, optical, embedded and other (force sensing) [24]. Among these, the

capacitive touch technology has been widely used and accepted. It has been employed

in mobile devices, consumer devices, as well as large-format displays. There are two

types of capacitive technologies in vogue, namely, surface capacitive and projected

capacitive. In a surface capacitive touch screen, a glass panel is covered by a transparent

7
electrode layer, which is enveloped by a protective cover [22]. When the screen is

touched by a bare finger or stylus, some of the electrical charge transfers from it to the

user. This leads to a decrease in capacitance which is detected by sensors located at the

four corners of the screen, allowing the controller to determine point of touch. This

technology is superior to the resistive one in terms of performance, and is relatively

insensitive to surface contaminants. However, it does not allow multi-touch. This

feature has been redeemed in the case of the projected capacitive technology [22].

Screens manufactured based on this improved technology can, in addition to an exposed

finger, be activated with surgical gloves or thin cotton gloves. It constitutes of a sheet

of glass with embedded transparent electrode films, and an IC chip, creating a three

dimensional electrostatic field. When a finger comes into contact with the electric field,

the computer is able to detect the points of touch using the change in the ratios of the

electric currents.

1.1.3.2 Ice layer detection

The presence of ice layers on the surface of runways and roads is a cause of concern

for the safe maneuvering of airplanes and automobiles. It makes continuous inspection

of the road surfaces mandatory in order to prevent freak accidents. Human inspections

may not be the best way to deal with such scenarios, owing to the frequency and

accuracy that these repeated checks demand. A sensor-based solution can be realized

in a non-contact manner using the capacitive sensing approach. This is based on the

relative permittivity of ice, and of water which depend upon the excitation frequency

given to the sensor and the temperature of the system [25], [26]. The relative

permittivity of ice is close to 80 when the applied excitation frequency is below 1 kHz.

It remains constant for a wide range of frequencies. As the excitation frequency is

increased beyond 20 kHz, the relative permittivity of ice will be around 3 [26].
8
An automated inspection of ice layers can be performed using a capacitive sensor that

is integrated into a remotely operated vehicle. This inspection vehicle can be moved

over the runways and roads with the ice layers [27]. The sensor is mounted to the bottom

part of the body of the vehicle. A multi-frequency excitation is employed to exploit

frequency-dependent property of the dielectric constant and detect presence of ice

layers reliably.

1.1.3.3 Capacitively-Coupled Parameter Sensing

Conductivity is an important parameter to determine the quality of water. The general

approach to determine water quality involves immersing a conductive probe in the

water to be tested. The critical requirement of contact in this resistive sensing method

makes it fraught with issues in reliability due to the deposition of organic and inorganic

compounds on the probe and subsequent corrosion. These issues can be avoided if a

non-contact, non-intrusive capacitively-coupled measurement approach is adopted.

Here the electrodes are coupled capacitively to the resistive water column as shown in

Fig. 1. 4.

Fig. 1.4 Non-contact conductivity measurement of water [28]

9
1.1.4 Interfacing Circuits for Capacitive Sensors

It is the interfacing circuit that extracts the measurand information that is embedded in

the signal from the sensing element and converts it into the form suitable for use by

systems for monitoring, automation and control. The utility, reliability and versatility

of a sensor depends heavily on the efficiency of the interfacing circuit since it needs to

faithfully capture and communicate the variations in the measurand for decision

making. Simple, yet sophisticated, interfacing circuits are invaluable as they reduce the

overall cost and power consumption of the system, while enhancing the performance.

Designing such circuits requires a clear understanding of the factors that may degrade

the sensor data, and an appropriate selection of the conversion paradigm to translate the

changes in the measurand.

One of the most common causes of concern while designing capacitive sensing schemes

is the presence of parasitic capacitance that can result in unwanted output [2]. Parasitic

capacitance to ground will be present when the shield of the cables, used to connect the

capacitive sensor to the interfacing circuit, is grounded. It can also exist between the

ground plane and traces in a printed circuit board or lines of a breadboard. Thus, it is

beneficial to develop a circuit which by virtue of its design tackles the sensitivity of the

output to the parasitic capacitance.

Another important consideration while designing capacitive measurement schemes is

the presence of offset capacitance. Generally, the capacitive transducers used in practice

produce very small changes of capacitances (usually, in the picofarad range) with the

changes in measurand. Thus, if not compensated for, the offset also gets converted, and

thus degrades the resolution of the capacitance measurement.

10
In the case of capacitive touch sensors, which require unshielded or partially shielded

sensing electrodes, a major challenge is the electrode sensitivity to electromagnetic

interference, especially power-line interference (both 50 Hz and 60 Hz). So far, the

methods used to deal with this issue at a circuit level are filtering techniques and

synchronous demodulation. However, there is a tradeoff between effectiveness and

circuit complexity in these methods [29]. The capacitance-to-digital converter (CDC)

that converts the sensed capacitance to a digital signal, either by first converting it to a

voltage and then using an ADC [2], or by modifying its design to give an easy-to-

digitize output without using a separate ADC, is predominantly used with touch

sensors. However, most of the available CDCs are highly susceptible to interference

[30]. Though some CDC designs such as the sigma-delta and dual-slope integrating

ones have tried to achieve interference rejection, they have had only limited success

[31], [32].

As discussed in the preceding sections, both the single-element as well as the

differential capacitive sensors are widely used. These sensors may be excited by either

dc or a sinusoidal ac source depending on the desired objective. Interfacing circuits that

are compatible with both single-element capacitive sensor and DCS adds to the

versatility of applications. However, such circuits are rare [33], [34]. Existing circuits

are mostly designed for either single-element sensors or DCS. There are some single-

element sensor interfacing schemes that can be extended to multiple sensors using

multiplexers [35]. An important feature to consider while developing an interface

scheme that works for both types of sensor is that the output is linear irrespective of the

sensor characteristic being linear or inverse. One available open-loop CDC [36] that

can interface both sensors fails to exhibit this feature, while additionally being

complicated in design, and sensitive to parasitic capacitance. Thus it will be useful to

DCS.

In the design of interfacing circuits that are compatible with both single-element and

DCS, which provide a linear digital output insensitive to parasitic and offset

capacitances, a quasi-digital approach has not been extensively explored. Interfacing

circuits that provide a quasi-digital output, like frequency or pulse-duration, are gaining

popularity due to their simplicity, low power consumption, high update rate sans high

complexity, the possibility of direct transmission over comparatively noisy or non-

linear channels and the facility to digitize these signals using digital counters [33], [37]-

[39]. Capacitance-to-frequency converters (CFCs) are relatively simpler and less power

demanding circuits than CDCs. In the CFC circuits the sensor capacitance is converted

into an equivalent frequency signal that can be easily measured using a counter.

However, when used in an open-loop configuration, their practical realizations are

negatively impacted by the high sensitivity to temperature and process parameters [37].

Typically, CFCs require a voltage-to-frequency converter (VFC), which leads to

increased complexity and cost of the interfacing circuit. Hence, the CFC which can be

an acceptable quasi-digital method when using a dc source and a closed-loop

configuration, may not be the best design for a measurement scheme that needs a

precise sinusoidal excitation. This calls for a simpler quasi-digital scheme, the

capacitance-to-pulse-width converter (CPC), which has all the attractive features of the

CFC [37]. The CPC can be easily interfaced to a digital system by synchronizing the

frequency of the pulse-width modulated (PWM) signal to that of the inbuilt clock of the

microcontroller [39] without the additional cost and complexity of additional analog

blocks like the VFC. A closed-loop configuration for the CPC is yet to be explored.

Also, no CPC has yet been designed for sensors that require sinusoidal excitation.
12
In the case of capacitively-coupled parameter sensing application of non-contact

resistance measurement discussed in section 1.1.3.3, the existing schemes for

capacitively-coupled resistive sensing [28], [40], [41] have their limitations. Generally,

these interface circuits employed are relatively complex with the requirements of a

precise sinusoidal excitation, and a phase-sensitive detector [28], [41]. The complexity

in these circuits is predominantly because they need to measure small resistances in the

range of tens of ohms (<100 Ω). However, in applications that involve the scaling of

the circuit for an array of sensors or the measurement of higher resistance values, such

levels of complexity are not required. In fact, this is the case for the typical sensor probe

dealing with drinking water which has its conductivity varying from

50- 800 S/cm [42], making the resistance to be measured be in the range of hundreds

of kilo-ohms. Thus, for the capacitively-coupled measurement of higher resistances, a

simpler scheme is preferred. One possible approach is the direct-microcontroller

interface that involves no active components between the sensor and the

microcontroller [43]. It is simple, low cost and consumes relatively less power.

However, it has not yet been employed in contactless resistive sensing. Such a simple

interface will be very promising in the water quality monitoring to determine the

potability of drinking water [42] and the quality of river water [44], [45].

1.2 MOTIVATION

From the above discussion it is evident that simple, yet efficient, capacitance

measurement circuits that can obtain the values of the parameters of interest from the

sensor, in the digital domain, with high linearity, accuracy and repeatability, need to be

developed. Also, converters that possess negligible sensitivity towards offset voltage,

bias current, parasitic capacitance, and other circuit non-idealities, are preferred by

designers, engineers and scientists.

The objectives of the research work conducted and presented in this thesis are listed

below.

1. Design an interference-insensitive CDC that shows negligible sensitivity to

interference, by virtue of design. It should use simple dc excitation.

2. Develop an auto-balancing quasi-digital capacitance measurement scheme that is

compatible with both single-element as well as differential capacitive sensors. It

should be excited by a dc source, and utilize a capacitance-to-frequency conversion

approach with closed-loop to obtain a linear output independent of the sensor

characteristic.

3. Develop an auto-balancing quasi-digital capacitance measurement scheme for both

single-element and differential capacitive sensor applications that require a

sinusoidal excitation. It must use the capacitance-to-pulse-width conversion

paradigm to give linear output independent of the sensor characteristic.

4. Design a simple, low-cost, low-power direct microcontroller interface for

capacitively-coupled resistive sensors that are modeled using a series RC model.

The series equivalent model of the capacitance and resistance better embodies the

capacitive sensors used in applications such as conductivity measurement of water

than a parallel model.

5. Ensure that in each of the designs proposed in objectives 1-4, the final output is in

digital form. The outputs in each design should be relatively insensitive to circuit

non-idealities such as parasitic capacitance of sensor, offset voltage, bias current,

and offset or nominal capacitance values in cases where the output is degraded by

14
 dependence on these factors. Further, the circuits must be characterized

experimentally by building appropriate prototypes and comparing their

performance with existing ones. In addition, performance optimization shall be

performed using circuit simulation and/or analysis before final prototype

development.

1.4 ORGANIZATION OF THE THESIS

Chapter 1 gives a brief introduction into the role of sensors in today’s world. Further it

elaborates on the sensor of interest, namely capacitive sensors. Some of the important

configurations in which these sensors are available are discussed. Certain capacitive

sensing applications that illustrate the significance of capacitive sensors are explained,

which lead to the motivation behind designing and developing advanced interfacing

circuits for these non-contact sensors. The objectives that were formulated ahead of the

development of circuits so as to realize simple, effective capacitive measurement

schemes, are elaborated in this chapter. The organization of the thesis is also presented.

Chapter 2 presents a new interference-insensitive switched capacitor (SC) capacitance-

to-digital converter (CDC) that is vital in capacitive sensing applications with open-

electrode. Its principle of operation, the impact of circuit non-idealities and noise on

the final digital output, as well as the development of the hardware prototype, and the

test results obtained through exhaustive experiments using it, are detailed in this

chapter.

Chapter 3 deals with a new closed-loop switched-capacitor capacitance-to-frequency

converter (CFC) which is compatible with both single-element and differential

capacitive sensors. This CFC uses dc excitation and provides a linear digital output,

irrespective of the sensor characteristic being linear or inverse. Additionally, the output
15
is insensitive to parasitic, nominal and offset capacitances. The operating principle of

the proposed CFC, along with the analysis of the effect of circuit non-idealities on its

performance are presented in this chapter 3. Also, the experimental studies based on

the developed prototype are detailed here.

In chapter 4, a novel auto-balancing capacitance-to-pulse-width converter (CPC) that

uses sinusoidal excitation, and operates in a closed-loop configuration is presented.

This capacitance measurement scheme is very useful in applications that need a

sinusoidal ac signal for excitation or those that utilize the frequency dependence of the

dielectric [27], [46]. Unlike most of the existing CPCs, the proposed interface circuit is

also compatible with both single-element and differential capacitive sensors. The final

digital output is independent of the nominal capacitance of the sensor and has a linear

characteristic irrespective of the sensor characteristic being linear or inverse. The CPC

is designed such that the PWM output depends on the change in the sensor capacitance

alone. It is insensitive to parasitic capacitance and has very low sensitivity to the non-

idealities of the components and ICs used. The operating principle, analysis of the

impact of circuit non-idealities on the output, and the experimental results from the

hardware prototype are included in this chapter.

Chapter 5 details a novel approach to directly interface a capacitively-coupled resistive

sensor to a microcontroller. It offers an attractive alternative for measurement of the

resistance of the sensing element, by eliminating the need for direct contact of

measurement circuit to the resistive element. The proposed direct-microcontroller

interface is simple, low-cost, and low-power. It becomes key in the measurement of

higher value of resistances, which are utilized to determine the quality of drinking

water. The design and working of the proffered interface are presented in this chapter.

16
The effect of circuit non-idealities are analyzed, and the performance evaluated using

hardware prototypes. The details of the same are elaborated in this chapter.

Each of the first-of-its-kind measurement schemes presented in chapters 2 to 5 are

simple in design, but efficient in performance. All of them are insensitive to parasitic

capacitance and provide digital output. Extensive analysis has been conducted on every

one of the proposed schemes to assess and quantify the effect of various error sources

on the final digital output, and offer design criteria than can make the effect, if present,

negligible. The real-world operation of each of the proposed circuits is tested by

building suitable prototypes, and conducting experiments to evaluate important

performance parameters. The results are presented in the corresponding chapters.

Further, these are compared with existing schemes used to perform the intended

measurement.

Chapter 6 outlines the summary, conclusion and scope for future work.

17