Electronic Control Systems With Acoustic Computer Science Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

The goal of this thesis was to make a comprehensive review of commonly used acoustic sensors and their accompanying electronic circuits. The paper is focused mainly on the sensors used in military applications, alarm technology as well as in various kinds of industrial processes. This thesis presents the examples of these sensors, the principle of their operations as well as their parameters and characteristics. The project was supported by the acoustic control system, which analyses the sound of the milk being mixed in the glass in order to control its level. The laboratory system is shown in the figure:

C:\Users\Usuario\Downloads\photo (12).JPG

It is composed of a directional microphone, simple milk frother, a tripod, a computer with appriopriate software (Audacity, MatLab) and a glass of milk. Measurements were conducted in series - for different positions of a frother in a glass. The next step was to analyze the recorded sound, both in the time domain and in the frequency domain. The results showed that, based on the sound analysis, it can be determined when the frother is on the correct position, or when it is too shallow or too deep. Therefore, in the further stages, the system can be expanded so that it could operate fully automatically.


Celem niniejszej pracy inżynierskiej byÅ‚o dokonanie kompleksowego przeglÄ…du powszechnie stosowanych czujników akustycznych i towarzyszÄ…cych im ukÅ‚adów elektronicznych. GÅ‚ówny nacisk padÅ‚ na czujniki stosowane w technice wojskowej (sonary), technice alarmowej a także w procesach przemysÅ‚owych. W pracy zostaÅ‚y przytoczone przykÅ‚ady tych czujników, zasada ich dziaÅ‚ania, a także ich parametry i charakterystyki. Projekt zostaÅ‚ poprarty akustycznym systemem kontroli, który analizuje dźwiÄ™k mleka mieszanego w szklance w celu kontroli jego poziomu. System ten widoczny jest na rysunku:

C:\Users\Usuario\Downloads\photo (12).JPG

SkÅ‚ada siÄ™ on z mikrofonu kierunkowego, spieniacza do mleka, statywu, komputera z odpowiednim oprogramowaniem (Audacity, MatLab) oraz szklanki z mlekiem. Pomiary zostaÅ‚y zrobione w seriach - dla różnych gÅ‚Ä™bokoÅ›ci zanurzenia mieszadÅ‚a. Kolejnym krokiem byÅ‚a analiza zarejestrowanego dźwiÄ™ku, zarówno w dziedzinie czasu jak i w dziedzinie czÄ™stotliwoÅ›ci. Wyniki pokazaÅ‚y, iż na podstawie analizy dźwiÄ™ku można stwierdzić, kiedy spieniacz znajduje siÄ™ na prawidÅ‚owej gÅ‚Ä™bokoÅ›ci bÄ…dź kiedy jest za pÅ‚ytko lub za gÅ‚Ä™boko. DziÄ™ki temu w dalszych etapach można rozbudować system tak, aby w przyszÅ‚oÅ›ci mógÅ‚ dziaÅ‚ać w peÅ‚ni automatycznie.

Table of Contents

Introduction 6

Sound and acoustics 7

Sensors 11

Acoustic sensors 12

In military 13

Sonar equations 15

Passive ranging 17

In alarm technology 19

In industry 24

Project of an electro-acoustic control system 28

Description 28

Purpose of the project 30

Laboratory system 31

Results 32

Summary and conclusions 40

References 42


The topic of my thesis - Electronic control systems with acoustics sensors is very broad and a little unclear. However it is not random. I wanted to combine detailed theoretical research on the topic that I am interested in with a practical project.

Sensors and control systems play crucial role in modern world and they are a way of the future. Obviously, their parameters are changing over time; however the fundamental principles of their operation and design remain the same. That is why I decided to make a comprehensive review of commonly used sensors (acoustic ones) and their accompanying electronic circuits - to show how many applications they have and how useful they are.

What is more, this paper is also about laboratory systems that measure and record acoustic waves in order to control something, for instance the level of the liquid in the container. In general, systems like this are composed of several elements. The first, and the most essential one, is a sensor, for example a microphone that responds to the acoustic waves (nonelectrical value) and converts them into a readable output (electrical value). Second, also very important, component is an electronic circuit, which powers the sensor as well as amplifies and filters analogue electrical signal from the microphone. Later on this signal can be left in analogue form or can be converted to the digital form. Obviously the second case gives much greater potential for further analysis of the recorded sound. Finally, this system can control something. It can turn on/off some things, but it can also control something on a continuous basis.

To support all the aforementioned theoretical background I decided to develop a model of such laboratory system to see if and how well it works. Moreover, I wanted to establish if, in further stages, it could be expanded so that it could operate fully automatically. This part was particularly interesting for me, due to the fact that for the first time since I have been studying I was able to develop the practical project entirely by myself.

Sound and acoustics

Since my thesis is mainly about acoustic sensors, it is necessary to explain the background of their operation. It is common to think that acoustics is the study of music in general. Although acoustics does comprise the study of musical instruments, it also includes an extensive range of topics, including: SONAR systems, noise control, ultrasounds for medical imaging and other processes, electroacoustic communication, seismology, etc. [1] In general, acoustics is the study of mechanical waves including sound, vibration, infrasound and ultrasound.

When talking about the acoustics, it is essential to answer the question: What is sound? It is nothing more than an acoustic wave produced when an object vibrates and communicates its motion to the surrounding medium. However, mechanical vibration does not have to cause the sound wave, due to the fact that a sound wave requires a medium that can be vibrated. Therefore, there is no transmission of sound in the vacuum. Opposite to solid media, such as wood iron or brick, a liquids and gases are not able to transmit transverse forces. That is why in this case sound waves are always longitudinal waves, which means that the particles move in the direction of propagation of the wave.[2] However, since this thesis deals only with propagation of sound in liquid and gaseous medium, I will focus mainly on longitudinal waves. They are composed of compressions, where the parts of the medium closer together than normal, and rarefactions, where the parts of the medium are farther apart than normal. [3]

Every acoustic wave is characterized by several parameters. The first one is frequency (f), which is the number of pressure variation cycles in the medium per unit time, or simply, the number of cycles per second. [4] The unit of frequency is Hertz [Hz]. The frequency range over which sound can be heard by the human ear is limited to the range of about 20 Hz to 20 kHz. Longitudinal mechanical waves below 20 Hz are called infrasound and above 20 kHz, they are called ultrasound. Waves with the frequency below 400 Hz are considered to be low, waves in the range from 400 Hz to 3 kHz as an average and those at frequencies above 3 kHz are considered high.

Second very important parameter is wavelength (λ), which is the distance travelled by the wave during one cycle. It is given by the equation: , where v is a speed of sound in the specific medium and f is aforementioned frequency of the wave.

A sound wave is also characterized by a sound pressure (p), which is the local pressure aberration from the ambient atmospheric pressure caused by a sound wave. It is measured in pascals [Pa]. There is also a sound level (SPL) expressed in decibels [dB] and it is a logarithmic measure of the effective sound pressure of a sound relative to a reference value. It is given by the equation: , where is the reference pressure. The graphical representation of a sound wave is presented on the following figure:

Sound pressure




Fig.1. Representation of a sound wave

Another essential parameter of an acoustic wave is sound intensity (I), which is the measure of the energy of an acoustic wave. Its unit is [W/m2]. It is equal to the mean value of the acoustic energy flux flowing during the time of 1 second by a unit area (1m2) oriented perpendicular to the direction of wave propagation. This value is rather difficult measure. Therefore it is usually expressed in terms of sound pressure: , where p is a sound pressure, v is a speed of sound and is a density of the medium. Obviously, the sound intensity is proportional to the square of the sound pressure. When talking about the sound intensity, it is important to mention about the sound intensity level, which is used very often. It is a logarithmic measure of the sound intensity in comparison to a reference level. It is given by the equation [5]: [dB].

Besides the parameters of the sound, there are also a few concepts related to this topic, which are essential in sound analysis. One of them is tone, which is the sound that can be recognized by its regularity of vibration. A simple tone has only one frequency, although its intensity may vary. A complex tone consists of two or more simple tones, called overtones. [6] By using Fourier transform every sound can be represented as a graph of the harmonic spectrum. The most significant influence on the reception of sound has basic tone, which is the greatest common divisor of the component tones.

When talking about the sound analysis it is also crucial to mention about the noise. In general, noise refers to any unwanted sound. However, in relation to acoustics, noise is a group of sounds that has the continuous spectrum. A special kind of noise is called a white noise, which is the noise in which all the component frequency have the same amplitude. As it will be presented in further sections, the concept of noise is essential as far as this paper is concerned.

Another very important phenomenon is rumble. It refers to a low frequency sound from the bearings inside a turntable. It is based on the fact that during the generation of two sounds of similar frequencies one may hear the sound of the frequency of the difference of these two. Due to this fact, it is easy to generate the sounds with low frequencies, for instance for 300 Hz and 335 Hz, one hears the sound of = 35 Hz.


Sensors play very important role in modern automation and control of systems, due to the fact that they can detect most physical phenomena. They are also a crucial part of my project. Therefore, I would like to elaborate on them in this section. First of all, it is important to answer the question: What are sensors? To put it simple, sensors are devices measuring (sensing) an environment and converting this information into readable output, which today would preferably by electronic, but which may as well be communicated using other processes, for instance acoustic or visual. This output, which is a signal, can later be used for making decisions about the operations being conducted. Therefore, sensors can be considered as translators of a generally nonelectrical value into an electrical value (a signal which can be channeled, amplified, and modified by electronic devices). What is also worth to mention is that sensors do not function by themselves - they always exist as a part of a larger system that can combine many other detectors, signal processors, memory devices, data recorders, etc.

Fig.1. Sensor block diagram

Power source


Power source

Input variable Data


Output signal (usually electrical or visual/acoustic)



Input variable from the physical world

The generic block diagram for a sensor presented in the above figure focusses the attention on the role of a sensor as a link among a control system and the physical world. As it was mentioned earlier, sensor converts some quantity from the physical world into (typically) electrical signal. Later on signal processor conducts one or more of numerous mathematical operations on the detected value, such as filtering, amplification or digitizing. The measured and processed value is transmited to other subsystems or directly to a human.

It is important to note that very few sensing methods provide a digital output directly, and most digital outputs are obtained by converting from analogue quantities. This indicates that analogue to digital conversion determines the limits of resolution more than the sensor itself.

Sensors can be classified into various criteria. In general, they can be divided into two broad classes depending on how they interact with the environment they are measuring. Based on this criterion, sensors are divided into passive and active. Active sensors add energy to the measurement environment as part of the measurement process. Good example of this type of sensors is radar, which measures the distance to some object by actively send out the radio to reflect off of some object and measure its range from a sensor. In this thesis, however, I would like to focus on the other type of sensors - the passive ones, which do not add energy as part of measurement process but may remove energy in their operation. In the next sections of this thesis I will provide some examples of this type of sensors.

Another way to divide sensors is to consider all of their properties, such as what they measure (stimulus), what their specifications are, what physical phenomenon they are sensitive to or what are their applications. Sensors can be chemical, thermal, optical, automotive, etc. However, as the title says, this thesis is about acoustic sensors and that is why I will describe this type of sensors in the following subchapter.

Acoustic sensors

Acoustic sensors are sensors, which have an acoustic wave as a detection mechanism. In these sensors, vibratory characteristics, such as phase velocity or the attenuation coefficient, are affected by the stimulus.


Output signal

Transduction of the input quantity to the acoustic wave device

Input quantity:

electrical, chemical, physical, etc.

Fig.2. Schematic diagram of an acoustic sensor

What is worth to mention, is that the frequency range over which sound can be heard by the human ear is limited to the range of about 20 Hz to 20 kHz. The acoustic sensors, which are the subject of this chapter, however, are not necessarily restrained to these frequency limits and some can be used with ultrasound or infrasound.

Acoustic sensors have a lot of commonly known applications in modern world. Currently, use of these sensors is broader than only detecting the sound. Applications range over measuring displacement, concentration of compounds, etc. I will describe few of these applications in the following sections.

Acoustic sensors in military

Acoustic sensors play enormous role in modern military technique. They have been used in Navy since World War I. This section will be, obviously, dedicated to SONAR (SOund NAvigation and Ranging) systems, the operation of which is based on the wave propagation between the receiver and the target. They make use of transmitted and reflected underwater sound waves in order to detect submerged objects or measure the distances underwater. Just like sensors in general, sonars are divided into two categories: active and passive. Active sonar systems provide their own source of sound and listen for echoes reflected from the target which they are aiming to detect. However, from the point of view of my thesis, passive sonar systems are more important and that is why this section will be dedicated to this type of sonars only.

Unlike active sonar systems, passive sonars do not emit its own signal. Their energy is initiated at a target and then propagated to a receiver. For tactic reasons, passive sonars are considered to be the most important sensors in modern submarines. They have been used in military since World War I, although their basic principles were understood already in sixteenth century. They guide most of the basic submarine water operations, such as search, detection and tracking.

Shipping noise

Sea state noise

Biological noise

Sonar signal processing

Submarine Radiated noise

Sonar receive beam

visual display

earphones C:\Users\Usuario\AppData\Local\Microsoft\Windows\Temporary Internet Files\Content.IE5\BX6Q1J1B\MC900337886[1].wmf

Fig.3. Schematic diagram of passive SONAR system

The above figure presents the operation of basic passive sonar system. It is visible that it listens without transmitting, which distinguishes it from active sonar systems.There exist few types of passive sonars, but the most typical ones are: passive noise sonar, passive ranging sonar and acoustic pulse surveillance sonar. One sonar system can be composed of various types of sonars. These sonars together gather some information, which can be, in general, classified into two main categories: characteristic information and positional information. Positional information reflects target position and motion, such as distance, velocity and course. On the other hand, characteristic information includes target type and identity.

Sonar equations

When talking about sonar systems, it is important to mention about sonar equations, which are a method of merging all elements of the sonar process. They were developed during World War II in order to estimate performance of sonar systems, for example their maximal range, as well as to support their design. Knowledge of sonar range is crucial in military operations for the purpose of planning suitable strategic tactics.

As far as passive sonars are concerned, the most fundamental equation is signal-to-noise power ratio:

Which id dB would be:

SNR = SL - TL - NL + DI

, where:

SNR = - signal-to-noise ratio (dB)

SL = - radiated signal (dB)

TL = - propagation or transmission loss (dB)

NL = - total noise (dB)

DI = - directivity index (dB)

The radiated signal (SL) could depends on the objective, but in the case of military applications, it is usually a submarine. This signal depends on various elements: the type of target, its speed, depth, acceleration, operating mode, the depression/elevation angle as well as the aspect (0 degrees for a bow of target and 90 degrees for the beam of the target). Passive sonar systems are designed in order to take advantage of radiated noise and distinguish it from interfering noises, such as array self-noise or ambient noise. The radiated noise can be of two types: narrowband or broadband. A broadband signal has acoustic energy over wide range frequencies, whereas a narrowband signal is composed of discrete tones. Furthermore, signals may be also categorized by their temporal characteristics. They can be continuous, intermittent, or transient in duration.

Most sonar systems are designed in order to concentrate the acoustic energy into a narrower beam with the purpose of improving their efficiency. This effect is accounted for in the sonar equations by the directivity index (DI), which is a measure of focusing.

The next important factor of this equation is transmission (or propagation) loss (TL). It is a loss in intensity between the point of interest and a reference point. Obviously, the intensity of an acoustic signal decreases with range, but this factor is also dependent on other parameters, such as frequency, surface loss or sound speed versus depth.

And the last component of the equation - total noise (NL) or background noise is a sound generated by everything except the target. In case of submarines, this noise is created by waves, fishes and other biological activity.

Another basic sonar equation expresses the difference between signal-to-noise ratio (SNR) at the output of the beamformer and the detection threshold (DT). This difference is called the signal excess and is given by the formula:

SE = SL − TL − NL + DI - DT

, where:

SE - signal excess (dB)

DT - detection threshold - a value that has a specified (usually 50%) probability of detection for a required probability of false alarm

By this definition, if the signal excess is greater than zero, the probability of detection is greater than 50 %.

Passive ranging

Basic passive sonar systems give no information about the range of a target. A signal may belong to a close, quiet target or a noisy, distant target. Therefore, there exist several techniques that passive sonar systems use to estimate the range of a target. The first technique is called triangulation and is based on simple trigonometry. The best way to explain this method is to show the practical example:

C:\Users\Usuario\AppData\Local\Microsoft\Windows\Temporary Internet Files\Content.IE5\BX6Q1J1B\MC900337886[1].wmf





S C:\Users\Usuario\AppData\Local\Microsoft\Windows\Temporary Internet Files\Content.IE5\BX6Q1J1B\MC900337886[1].wmf

Fig. Ranging by triangulation

A submarine is in contact with the target at range R from the flank array and from a towed array whose centre is S = 1000m behind the flank array. From simple trigonometry we have:

This estimate is highly dependent on the accuracies of the bearing measurements.

Another method is called Vertical Direct Passive Ranging (VDPR) and is based on determining the vertical arrival angles of signals arriving at the same array through multiple paths as well as measuring the time differences between them. This method is presented on the following figure:




α C:\Users\Usuario\AppData\Local\Microsoft\Windows\Temporary Internet Files\Content.IE5\BX6Q1J1B\MC900337886[1].wmf C:\Users\Usuario\AppData\Local\Microsoft\Windows\Temporary Internet Files\Content.IE5\BX6Q1J1B\MC900337886[1].wmf




Fig. Ranging by VDPR method

To calculate the range between the platform and the target it is necessary to make some assumptions. The first one is that platform and target depths (p and t) have to be much less than the depth of water (H). And the other one is that the sea bed should be flat at the bounce side.

The equation for the range of the target in terms of arrival angle α is as follows:

for bottom bounce and surface bounce (BS) and:

for bottom bounce only (B).

The last method is based on the measurement of wavefront curvature using three well-separated arrays and is called Horizontal Direct Passive Ranging (HDPR). However, due to the fact that this method is used very rarely, I will not describe it in detail.

Acoustic sensors in alarm technology

In general, alarm system is a set of devices designed to detect intrusion into a building or area. Alarm systems have specific purposes. They can be used as fire alarms, pressure alarms and process alarms in addition to intrusion detection sensors. However in this paper I will focus only on alarm systems which operation is based on acoustic sensors. They are called glass break detectors (sensors) and usually are very simple in their operation.

Acoustic glass break alarm sensors make use of an omnidirectional microphone, which receives the sound of shattering glass. They can work with any security alarm control system. What is worth mentioning, they may also detect impulsive shockwaves of a breaking object, for instance a window. Glass break detectors monitor any vibration or noise coming from the glass. If the vibrations surpass some threshold (which sometimes can be selected by the user), they are then analysed by detector circuitry. Some detectors solely use narrowband microphones adjusted to frequencies characteristic for glass breakage, and react to sound above specific threshold, while more complex designs compare the sound analysis to one or more glass break profiles using signal transforms similar to FFT and react if both the statistically expressed similarity threshold and the amplitude threshold are exceeded.

A glass break event consists of an initial signal portion frequently denoted as a "thud", which is related to the initial impact between the glass surface and the striking object, pursued by the formation of cracks in the glass and followed by the damage of the glass. Subsequently, the glass fragments persist to resonate and hit other glass pieces as they reach the floor and/or surroundings. This latter portion is frequently referred to as a "tinkle" portion.

There exist two types of sensors with the purpose of detecting breaking glass: active and passive. Active detectors operate on the basis of detection of the acoustic pulse in response to the noise created when glass is broken, whereas passive ones are designed to respond to mechanical vibrations. Acoustic sensors are also divided into those that respond to high-frequency signals, for example crack of the window and those that detect the hit. They react to signals in the frequency band from 6 kHz to 30 kHz.

Acoustic glass break sensors depend greatly on the quality of sound events, due to the mode of their operation, causing many challenges to their designers. These detectors have to also be able to reject all failure alerts - sounds coming not from true glass shattering.

In general, glass break detectors have to always be "on" and should be able to handle any sound activity in real time. Nevertheless, user can turn off some of the sensor's blocks or set them into low-power modes while they are not working.

Glass break sensors can be designed with three types of implementations: wired connectivity, wireless connectivity and standalone. The last two of them are battery operated. The block diagram of the glass break sensor consists of three sub-blocks: the sensing block, the processor and the communication interface blocks. It is presented on the following figure:


Fig. Sample block diagram of a glass break detector.

There are various types of glass break detectors on the market. They differ with complexity, price and additional features, such as passive infrared motion detector or Pattern Recognition Technology. As the selection of these types of alarm sensors is very large, I decided to choose two most representative examples and describe them a little more thoroughly.

The first one is round flush mount glass break detector from Honeywell. It uses field proven technology to provide increased false alarm immunity and faster response. It uses the FlexCore Signal Processor, which is an Application-Specific Integrated Circuit (ASIC). It processes sound data in parallel in order to make more accurate and faster detection decisions. This detector can be mounted anywhere and its maximum range is 7,6m. It performs Multiple Domain Signal Analysis during which frequency, time and amplitude characteristics are evaluated for signal qualification. This feature allows the sensor to precisely distinguish true glass break events from false alarms.

Fig. Round flush mount glass break detector

Second example is from GE Security. It is called ShatterPro Plus and apart from being a glass break detector, it is also joined with a passive infrared motion detector (PIR) in order to eradicate occupant-generated false alarms.  This sensor also has a simple hand clap feature that allows the user to check if a handheld tester can be activated at close range or if the detector is operational. By processing over thirty frequency points across the glass break frequency spectrum, this detector is able to reject most common false alarm sounds, whilst offering good glass break detection. Obviously, this glass break sensor has more additional features than the previous one. However it is bigger and has smaller range - only 4,6m. It is visible on the figure below:


Fig. ShatterPro Plus



max: f =161.2022Hz y= 2.9532*10^7


max: f= 160.3211Hz y= 4.2863*10^7


max: f= 90.2707Hz y= 3.6184*10^7



max: f =145.5621Hz y= 7.5167*10^7


max: f= 144.9012Hz y= 9.4621*10^7


max: f= 133.6667Hz y= 10.781*10^7



max: f=124.6350Hz y=8.148*10^7


max: f= 65.5989Hz y= 13.941*10^7


max: f= 126.6176Hz y= 14.002*10^7


max: f= 122.2119Hz y= 10.333*10^7


max: f= 79.6970Hz y= 7.8921*10^7



max: f= 74.1899Hz y= 13.716*10^7


max: f= 78.5956Hz y= 21.487*10^7


max: f= 75.0711Hz y= 17.911*10^7



max: f= 63.6163Hz y= 28.306*10^7


max: f= 69.5640Hz y= 20.027*10^7


max: f= 68.2423Hz y= 33.191*10^7