This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
Noise cancellation is all about removing unwanted noise from surrounding environment while keeping the source sound as it is. Prime objective of this system is to make user's listening experience better in noisy environment. It has application in manufacturing industries, consumer products, and music equipment. This report includes study of feedback active noise cancellation to remove noise, what it is and how does it do noise cancellation. The report also consist various tests outcomes that has been conducted on developed prototype.
I am highly indebted to Chris Daykin for his guidance and constant supervision as well as for providing necessary information regarding the project and support during development of the project.
I would like to express my gratitude towards Professor Peter Mather, My parents and staff of T5/02 Laboratory for their kind co-operation and encouragement which help me in completion of this project, and my thanks and appreciations also goes to my fellow classmates and people who have willingly helped me out with their abilities.
Table of Contents
[Chapter - 1] Introduction, Aims and Management of the Project
Noise is aperiodic (non-periodic) signal because the frequency content, amplitude, phase and sound velocity of a noise signal are nonstationary (Kuo & Morgan, 1996).Noise cancellation is used where noise is significant enough to make annoying effects or can damage sensitivity of ear. This report concentrates on ANC as an application of headphone; external noise is reduced using ANC so music can be heard clearly. Generally all kind of headphones do reduces at least small amount of noise passively, due to the materials they were built from does not allow external noise to pass through, and limits it before it can be heard by ears (Kuo & Morgan, 1996)
"Circumaural (closed-back) headphones" are known to be the best for reducing noise passively (Harris, n.d.), because it closes ear completely with its sound absorbent paddings, since the passive noise cancellation is not really efficient at lower frequencies, ANC is used in headphone to attenuate lower frequency noise and together they provides noise reduction at wide range of frequencies (Kuo & Morgan, 1996).
This project also focuses various other possible configurations which can be implement and led to significant improvement on noise cancelling ability of system, such as alteration in placement of microphones.
The main Aim is to study Active Noise Cancellation and expand the gained knowledge by developing a prototype of the topic, which can be used further in real life applications (e.g. noise cancelling headphones).
1. Study Active Noise Cancellation and its various other configuration, and how it can be used in headphones to cancel ambient noise.
2. Fabricate only left channel system, test it and modify it again and again by changing circuit configurations to get best possible results.
3. Construct the right channel and combine it with left one, analyse outcomes of it as a stereo circuit by testing it with a headphone and auxiliary audio signal.
[1.4] Management of Project
To make the project developing operations easier and to meet academic deadline, the entire process of development was divided in to six main tasks, and each of the tasks was given a specific time limit which ends before the academic deadline ends, so in case if anything goes wrong, I shall still have sufficient time to fix it without taking any stress, regular time allocation to prioritizing the tasks that have higher importance is the key to make a smooth approach towards the deadlines.
(Figure 1-2: Project Schedule)
(Table 1-3: Days allocated to individual task with deadlines)
Task 1 - Study and collect information about Active Noise Cancellation.
Task 2 - Construct one channel of circuit test it and get the result out of it.
Task 3 - introduce various changes in circuit and notice its effect on performance.
Task 4 - Simulate ANC system using software.
Task 5 - Build full ANC circuit and test it with headphones.
Task 6 - Use gathered results to make comparisons and start to write final report.
Task 7 - Find other ways to improve performance.
All the laboratory work, experiments and test out comes will be carefully noted down to a logbook so, it can be used later while writing report and to make comparisons.
[Chapter-2] Literature Review
[2.1] Theory of Operational Amplifier
Theory behind all of three operational amplifier configurations that were used in this project is explained in this section of report.
An Op-Amp with an input signal connected to its non-inverting input is called a non-inverting amplifier (figure 2-1). Output of this op-amp is connected back to non-inverting input via potential divider, and ratio of resistors R1 and RF in potential divider determines the gain for this op-amp. The output from op-amp drives current into RF until negative input reaches at same voltage as input voltage (Mancini & Carter, 2009).
(Figure 2-1: Non-Inverting Amplifier Configuration)
Voltage divider rule is used to calculate output voltage.
Equation shown above can be rewritten as below to get the gain.
Inverting Operational Amplifier
As its name suggests in an Inverting op-amp input signal is applied to inverting input, and non-inverting input is left grounded so, the provided feedback via RF keeps inverting the input applied to op-amp (Mancini & Carter, 2009). The gain is set by ratio of RF and R1 resistors.
(Figure 2-2: Inverting Amplifier Configuration)
Kirchhoff's law is used to get the equation shown below.
It can be further simplified to get the gain of the inverting op-amp.
Summing Operational Amplifier
Summing (Adder) op-amp is just an inverting amplifier with more input (figure 2-3), feedback is provided in same was as inverting op-amp, but the feature of a summing amplifier is adding additional input signals does not change the response of the current inputs (Mancini & Carter, 2009).
(Figure 2-3: Summing Amplifier Configuration)
The output for individual input is calculated using superposition.
The total output voltage can be determined by adding above two expressions algebraically.
And if R1=R2 the equation can be further simplified to
[2.2] Noise Cancellation Theory
There are two main ways passive and Active in which noise cancellation can be done. Passive noise cancellation is most conventional and was only known way of noise reduction till the introduction of active noise cancellation (Kuo & Morgan, 1996).
Active noise cancellation does the job so well where passive noise cancellation is not really very efficient, at lower frequencies (Kuo & Morgan, 1996). ANC uses a system based on "superposition principle", it creates a secondary signal from primary noise signal but out of phase and with same amplitude and combines it with primary noise signal, as a result both waves cancels each other out and ends up in a waveform similar to waveform shown in (fig 2-4), the amount of noise attenuation depends on accuracy of amplitude and phase of created secondary secondary signal (Kuo & Morgan, 1996).
(Figure 2-4: Results of combining the waves of different phases - Image courtesy: mediacollege.com)
[2.3] Circuit Functioning
Because this is a stereo circuit it processes left and right signals separately but in same manner, so only one channel will be explained in this report however the tests that has to be carried out on the full circuit will be described in detail.
Electret type of Microphone has internal FETs, it need 2 to 10 bias voltages in order to operate that is provided by C1 bypass capacitor and R1 through R3 which formats a voltage dividing network. C2/R4 and C4/R6 are high pass filters; it blocks any DC before first op-amp stage set as a non-inverting amplifier. Ground reference is provided by R4 with high resistance value. R8 and R6 are chosen to give a gain of 31db for the non-inverting amplifier stage. Second op-amp stage is a phase Invertor with unity gain determined by ratio of R10 and R11 (Ryckebusch, September 1997).
(Figure 2-5: Functional Circuit Diagram of Active Noise Cancellation)
The last op-amp stage is called a Phase Invertor with gain established by ratio of R15 and R19, the setup of this of amp can easily be reformed from a Phase Invertor to Summing amplifier by adding R17 of and applying source signal from auxiliary input J2 to it, another intention to make the values of R15 and R17 same is to enable them to work with R14-a potentiometer logarithmically (Ryckebusch, September 1997).
(Fig 2-6: Various Operational Amplifier Stages)
The circuit has combination of three different op-amp configurations the first is non-inverting amplifier (IC1-a) with the gain of 31db, the signal comes from the microphone attached to the (J1) microphone in jack gets amplified by this stage of op-amp, output of the stage will be an amplified version of the microphone signal that makes this signal easy to be processed by next op-amp stages. Output of this stage is goes to two places; one is the second stage of op-amp set as a Phase Inverter (IC2-a), and the second place is first pole of (S1-a) switch.
Second stage simply inverts the incoming signal by with the gain of unity, resulted signal will be applied to second pole of (S1-a) switch. And switch now becomes a phase selector switch which can select out of phase inverted signal or the signal without any inversion.
The selected signal by (s1-a) switch then goes to (R14-a) potentiometer, after that signal reaches at the last op-amp stage a Phase Invertor (IC3-a) with the gain of 20db set by the ratio of R19 and R15. Setup of last op-amp can easily be reformed to a summing amplifier configuration by applying a source signal in from (J2) auxiliary in jack via (R21-a) potentiometer and R17 resistor, summing amplifier will cancel microphone signal with its inverted version and the only source signal will be left at output ready to be collected through overload protection resistor (R21) from (J3) headphone out jack.
(R14) potentiometer became vital to attenuate the microphone signal (noise signal) so its amplitude can be matched with the signal at output.
To sum up, the signal from the microphone is first amplified and then it will be inverted, after that summing amplifier will add microphone (noise) signal with its inverted version and they both will destruct each other out and the noise will be subdued.
[2.4] Accuracy of Active Noise Cancelling System
As it is necessary for secondary signal to be of same amplitude and exactly out of phase to primary noise signal for maximum noise attenuation, the time taken by ANC system to process secondary noise signal causes the problem of timing delay, due to noise signal arrives to ear before processed signal by ANC system can. That's why switch (S1-a) is added to address this issue it have the option to eliminate inverting op-amp stage when the delay is significant enough to make whole system unstable.
[2.4.1] Open and Closed loop ANC systems
Another approach that can be made in order to reduce timing delay is a closed loop ANC system in which microphone is placed inside of ear cup of headphone (Fig 2-7); in a closed loop ANC system microphone will receive noise and audio signal together emitted by transducer, then the first stage of close loop ANC extracts noise signal and removes the desired audio signal out of signal received by microphone, and after inversion additional noise will be filtered out by band pass filter, inverted signal then will be combined with desired audio signal and emitted via transducer (Moy, 2001).
(Figure 2-7: Open Loop and Close Loop ANC Systems - Image Courtesy (Moy, 2001) )
To minimize timing delay microphone is placed as close as possible to the headphone transducer, benefit of using negative feedback closed loop ANC system is higher amount of noise attenuation is possible almost of 24 - 30 dbs If good quality headphone with passive noise reduction were used, while in open loop ANC system attenuation is limited to 10 -14 dbs with same headphone (Moy, 2001).
The reason behind not using closed loop ANC system in this project is its issue with stability because of negative feedback it becomes far more sensitive to timing delays then open loop ANC system.
[2.4.2] Digital Implementation of ANC System
Any ANC system must match the amplitude and phase of primary noise signal and secondary signal accurately, and to do that it should have extremely precise control, stability and reliability to attenuate greater amount of noise (Kuo & Morgan, 1996), which is less desirable in an analogue system therefore it is more desirable for ANC system to be digital. Software called MATLAB is used to simulate digital ANC system using Least Mean Square (LMS) algorithm (fig 2-8), this adaptive systems constantly updates its coefficients using LMS algorithm (Kuo & Morgan, 1996).
(Figure: 2-8 Block diagram of Adaptive LMS ANC System - Image courtesy (Kuo & Morgan, 1996) )
This adaptive ANC system takes two input signal x(n) which is a noise signal and d(n) is desired audio signal and generates e(n) as combination of desired audio signal and cancelled noise signal.