Electronic Circuit Design (EE-313) DE-42 Mechatronics Syndicate – A Title: Complex Engineering Problem Names of Group Members as per CMS: 1. NC Aliza Naeem 2. NC Maryam Awan Reg # 337922 Reg # 347962 Submitted to: Assistant Professor Kanwal Naveed Problem Statement: A stereo noise creates a huge disturbance, especially when you are listening to weak radio stations. Peaks of unwanted background noise take over the broadcast signal, making it unpleasant resulting in a low Signal to Noise Ratio (SNR). This attenuation might cause the signal to be inaudible most of the time. Therefore, the application of an appropriate amplifier is essential. We have to make a filter to remove noise and have to use an amplifier to amplify the noise free signal. Solution: Designing of filter: We have used a gaussian filter in MATLAB to remove the noise from our input signal. A Gaussian filter is a filter whose impulse response is a Gaussian function (or an approximation of it since the true Gaussian response has an infinite impulse response). Gaussian filters have the property of having no ringing on the step function input while minimizing rise and fall times. This behavior is closely related to the fact that Gaussian filters have the lowest possible group delay. Gaussian filters have the best combination of high frequency rejection while minimizing spatial dispersion. This is the key point of the uncertainty principle. INPUT and OUTPUTS: Figure 1 Original Signal Figure 3 Gaussian low pass filter Figure 2 Noisy Signal Figure 4 Signal after noise removal Design of Butterworth 3rd Order Filter: A Butterworth filter is a type of signal processing filter designed to have a passband response that is as flat as possible. Butterworth had a reputation for solving "impossible" math problems. At the time, filter design required a significant amount of designer experience due to the limitations of the theory in use at the time. This filter was out of common use for over 30 years after its introduction. Butterworth explained: “An ideal electrical filter should not only completely eliminate unwanted frequencies, but should also have uniform sensitivity to desired frequencies.” Although such an ideal filter cannot be achieved, Butterworth showed that as the number of filter elements with correct values increases, a closer approximation is obtained. At the time, filters produced large passband ripples and component value selection was highly interactive. Butterworth showed that it is possible to design a lowpass filter whose cutoff frequency is normalized to 1 radian per second and whose frequency response (gain) is Where ɷ is the angular frequency in radians per second and is the number of poles in the filter— equal to the number of reactive elements in a passive filter. If ɷ = 1, the amplitude response of this type of filter in the passband is 1/√2 ≈ 0.7071, which is half power or −3 dB. This was our attempt to design a filter. Implementing the electronics part. Proteus Simulations: Figure 5 Proteus Implementation Figure 6 Output Audio Figure 7 Input Audio Amplifier Design: Figure 8 Proteus Implementation Figure 9 Input Output ANALYSIS: We choose a two stage BJT amplifier in common emitter configuration. The audio file we chose was “forest-3-26722”. The input signal was very low in magnitude. The filter we designed was a third order butter-worth filter. Understanding SNR: The SNR understanding was done in MATLAB. Where we generated our own noise signal with a particular SNR. We simulated a sine wave and noise signal. The filter designed in MATLAB was 3rd order butterworth Filter to understand the SNR and noise better. Code: clc; close all; alphap=4; alphas=30; fp=1200; fs=1000; f=8000; ohmp=(2*pi*fp)/f; ohms=(2*pi*fs)/f; [N,wn]=buttord(ohmp,ohms,alphap,alphas); [B,A]=butter(N,wn,"low"); t=0:0.01:pi; [h,ohm]=freqz(B,A,t); subplot(3,1,1) plot(ohm/pi,abs(h)) grid on; xlabel('normalized frequency') ylabel('gain') title('frequency response') subplot(3,1,2) plot(ohm/pi,20*log(abs(h))) grid on; xlabel('normalized frequency') ylabel('gain in db') title('frequency response in db') subplot(3,1,3) plot(ohm/pi,angle(h)) grid on; xlabel('normalized frequency') ylabel('gain in db') title('frequency response in db') Amplifier Cascaded Two Stage BJT Amplifier: The calculations of Amplifier were done with following assumptions: Vcc = 12V, VE = 10% of VCC. 2NPN BJTs were considered for design. BC547 and 2N3904. Cut-off Frequency = 1kHz. Input Signal => Vp-p = 200mV Av for first stage is a little less than second stage. We acquired our Quiescent point by midpoint biasing. Dc gain: β = 100. Using Equations of Common emitter, we calculated the values of our resistors and capacitances with cut off frequency. Higher cut-off frequency 44.68 M Hz was calculated with miller effects.