Suman Chalise Roll No. 32014 EE 3rd year. Signal processing Tasks and Factors to be considered: 1) Conversion of signal CT DT Analog Digital Mathematical/ Theoretical CT to DT: Sampling Impulse Train Sampling Mathematically it is hard to do sampling thus we perform Impulse Train Sampling. In impulse we manipulate x(t) with impulse train s(t). x(t) y(t)=x(t)*s(t) s(t) X(j) × Y(j) = 1/2 *X (j)×S(j) S(j) -Convolution with impulse gives same function. -Convolution with shifted impulse gives shifted function. c , maximum practical frequency containing of signal. s ,sampling frequency. We can preserve the signal after sampling if c ≤ s / 2, s ≥ 2c. this is the Sampling Theorem (Nyquist Criteria). If s < 2c, then there is spectral corruption after sampling, this is known as Aliasing. If there is aliasing, the original signal cannot be recovered from sampled signal. Analytically Y (j) is given by, +∞ Y(j)= 1/T X [ j(-ks)] -∞ H(j) = 1 for ≤ s/2 0 , otherwise Y (j) × H (j) = X (j) X (j): we are unsure of exact value of c. First before sampling, we need to reduce the effective bandwidth of X (j) to maximum s/2. This may reduce the signal corruption. This is called using the anti-aliasing filter. Anti-aliasing filter is compulsory in every sampling system (ADC). xD [n ADC x(t) y(t) s(t) 4-bit unsigned Conversion to digital xD [n] = {0,1,2,…..,15 Quantization noise added. y(t) S/H t ADC is performed to scale the signal to fit into dynamic range of ADC. Due to digital signal conversion there is quantization noise added. Noise power =∆2 /12 ; ∆-step size (min.) This is normally called minimum noise, if signal fits exactly to the dynamic range of ADC. # Digital to Analog Conversion. DAC XD[n] yr (t) Digital S/H waveform-(Sample and Hold waveform) Comparison between Analog and Digital filters Analog filter In analog filter both inputs and outputs are continuous-time signal. Implementations of such filter are carried out using passive components. Analog filter operates in infinite frequencies range. Main disadvantages of this is, higher noise sensitivity. Digital In digital filters both inputs and outputs are discrete-time signals This filter is implemented on digital computer or DSP. Here frequencies are limited to half of sampling rate. It requires additional A/D and D/A converter. 2) Filtering -Filtering is for noise and inference reduction. -Selection -Equalization -Recovery 3) Modulation AM FM PM PWM 3) Multiplexing