# Q1 what is the difference between butterworth , chebyshev and elliptic filter..????

The Butterworth Filter is the maximally flat amplitude filter. It provides a near 0 attenuation until near the cutoff frequency and then descends into attenuation smoothly. The tranition becomes sharper with higher orders. It has moderate group delay so it has some overshoot on sharp rising waveforms. this gets worse with higher orders. The Chebyshev filter trades off flatness in the pass band for a steeper decline into the stop band. You design a cheby with a recurring wavelike ripple of attenuation in the paasband of usually 0.05db to 3db. In return you get a much steeper portion of the attenuation curve near the cutoff frequency. waveforms are distorted by group delay errors more severely than in the butterworth. The higher the ripple the worse the distortion The elliptical filter is like cheby^2 since it has ripples in the stopband and the passband. It has been proven to be the fastest possible descent into the stopband. The time delay errors are more severe than the cheby. For all of these filters. when at a distance from the cutoff frequency they will attenuate 20db/decade in frequency times the filter order.

Q2 what is 1) unit impulse signal 2) unit step signal 3) ramp signal
1) The unit impulse is a signal which is zero everywhere except when its argument is zero, then it is equal to 1. Mathematically

(n)={1 if n=0 0 otherwise
The unit sample function is very useful in that it can be seen as the elementary constituent in any discrete signal. Let

x(n)

be a sequence. Then

see Figure 1 for plots. In NMR an exponentially-shaped free induction decay (FID) signal is acquired in the time domain and Fourier-transformed to a Lorentzian line-shape in the frequency domain. The Fourier transform is compatible with differentiation in the following sense: if f(x) is a differentiable function with Fourier transform .we can express x(n) as follows (using the unit sample definition and the delay operation) 2) The unit step function is equal to zero when its index is negative and equal to one for non-negative indexes. This can be used to transform differential equations into algebraic equations. e. UNIT STEP: u(n)={1 if n 0 0 otherwise 3) Ramp signal 3) Ramp signal The ramp function. denoted by r ( t ) is a signal whose amplitude increases proportionally as time increases. By extending the Fourier transform to functions of several variables partial differential equations with domain Rn can also be translated into algebraic equations. The Fourier transform is also used in magnetic resonance imaging (MRI) and mass spectrometry. . then the Fourier transform of its derivative is given by. . infrared (FTIR). Fourier transform spectroscopy The Fourier transform is also used in nuclear magnetic resonance (NMR) and in other kinds of spectroscopy.shows a ramp signal of the function A discrete-time form a ramp signal is called a ramp sequence and is shown in Fig. The mathematical definition of ramp signal is Fig. Note that this technique only applies to problems whose domain is the whole set of real numbers.Q3 practical application of fourier transform?????? Analysis of differential equations Fourier transforms and the closely related Laplace transforms are widely used in solving differential equations.g.

As such. which makes it easy to see what frequencies the signal contains in order to filter/manipulate particular frequency components. Q6 difference b/w message and information(weblogs)????? Below I make assumptions and generalizations about message board and weblog design. which is also a bilateral Laplace transform evaluated at s = iw. both domains are continuous and unbounded. In this specific case. In other words. In mathematics. The term Fourier transform can refer to either the frequency domain representation of a function or to the process/formula that "transforms" one function into the other. I realize that the . representing signals in this way allows one to see the harmonics in a signal distinctly. whereas a DFT takes approximately N^2 operations. so the FFT is significantly faster simple answer is FFT = Fast DFT Q5 significance of fourier series & transform???? The Fourier series is important because it allows one to model periodic signals as a sum of distinct harmonic components. My goal is to discuss what I think are standard practices across the technologies. it transforms one function into another. the continuous Fourier transform is one of the specific forms of Fourier analysis. An FFT (Fast Fourier Transform) is a faster version of the DFT that can be applied when the number of samples in the signal is a power of two. which is called the frequency domain representation of the original function (which is often a function in the time-domain).Q4 Difference b/w FFT & DFT??? A Discrete Fourier Transform is simply the name given to the Fourier Transform when it is applied to digital (discrete) rather than an analog (continuous) signal. An FFT computation takes approximately N * log2(N) operations.

To make this more concrete.assumptions below may or may not match with your experiences and I present them as suggestions. equal in length). however implicitly it is used very frequently. My conclusion is that online communities will use the two resources to fill two different roles. and sum while shifting one function with respect to the other. Explicit circular convolution is rarely used. Below the table is a description of each row. the convolved output is also periodic and so the convolved output is fully specified by one of its periods. or in discrete time using vectors. let me leave you with a simple example (in pseudo-MATLAB notation). I think of it as flip. x = [1 2 2]. it is as if the finite length functions repeat in time. maybe always. Here I will use the FFT to compute the circular convolution. periodically. Any time DFTs (FFT) or Fourier Series are multiplied. there is an underlying circular convolution taking place. It also holds for functions defined from -Inf to Inf or for functions with a finite length in time. In circular convolution. Basically it is a correlation of one function with the time-reversed version of the other function. I believe that weblogs and message boards *are* different -different enough to happily exist together in the same online community web site. which I will call time. y = [1 3 1]. Because the input functions are now periodic. multiply. % Notice the result has tails where x and y do not fully overlap X = fft(x).y) == [1 5 9 8 2]. conv(x. where the convolution sum is an integral. Q7 difference between linear and circular convolution????? Linear convolution takes two functions of an independent variable. Circular convolution is only defined for finite length functions (usually. Their ability to fill independent niches will make the subtle differences between them make more sense. The table below outlines the differences I see. Please comment or email me with any input. This holds in continuous time. . First. and convolves them using the convolution sum formula you might find in a linear sytems or digital signal processing book. where the sum is truly a sum. continuous or discrete in time.

whereas the linear convolution is finite% length. the procedure is known as crosscorrelation. then the % circular convolution is the same as the original linear convolution. then the correlation coefficient is -1. and when the same signal is compared to phase shifted copies of itself. or any other dimension. % No tails and the result is different because x and y are assumed periodic % Now if we zero pad x and y to avoid the overlap between one period and the next. Y = fft(y). the distance between correlation peaks is taken to be the fundamental period of the signal (directly related to the fundamental frequency). For signals that vary with time. ifft(X*Y) == [9 7 9]. is the average magnitude difference function. If the signals are identical. X = fft(x). time. or with other low-pass filter operations. if they are totally different. the correlation coefficient is 0. and the degree of correlation in terms of frequency and phase represents the frequency and phase spectrums of the input signal. y0 = [1 3 1 0 0]. ifft(X*Y) == [1 5 9 8 2]. Q9 what is sampling theorem & give nyquest criterion ?????? Sampling can be done for signals varying in space. and if they are identical except that the phase is shifted by exactly 180degree (i. the sampled signal x[n] given by: . Though strictly speaking % the circular convolution is infinite-length and periodic. Q8 difference between auto & cross correlation???? Correlation determines the degree of similarity between two signals. The method may be combined with the simple smoothing operations of peak and centre clipping. Autocorrelation is a method which is frequently used for the extraction of fundamental frequency. Cross-correlation is the method which basically underlies implementations of the Fourier transformation: signals of varying frequency and phase are correlated with the input signal. and similar results are obtained in two or more dimensions. let x(t) be a continuous signal to be sampled. then the correlation coefficient is 1. which is called the sampling interval. :if a copy of the signal is shifted in phase. A function which is related to the correlation function. When two independent signals are compared. the procedure is known as autocorrelation.Y = fft(y). and let sampling be performed by measuring the value of the continuous signal every T seconds.e. Thus. but arithmetically less complex. mirrored). x0 = [1 2 2 0 0].

The sampling frequency or sampling rate fs is defined as the number of samples obtained in one second. . a frequency component with frequency f cannot be distinguished from other components with frequencies NfN + f and NfN ± f for nonzero integers N. if the sampling rate is more than twice the maximum frequency. 1.. of course. The sampling rate is measured in hertz or in samples per second. This ambiguity is called aliasing. Reconstruction in this case can be achieved using the Whittaker±Shannon interpolation formula. That is. necessary in practice). The frequency equal to one-half of the sampling rate is therefore a bound on the highest frequency that can be unambiguously represented by the sampled signal. The sampling theorem guarantees that bandlimited signals (i. used in creating each output sample is called the order of the filter. or fs = 1/T.represents the class of all finiteorder causal LTI digital filters. signals which have a maximum frequency) can be reconstructed perfectly from their sampled version. If and in Eq. in samples. but their frequency is ambiguous.. For example. Frequencies above the Nyquist frequency fN can be observed in the sampled signal. are constrained to be finite (which is.e. 2.. specifies a particular second-order filter. 3. Q10 what do you mean by order of filter??? Filter Order The maximum delay. In the difference-equation representation.. with n = 0. then Eq. This frequency (half the sampling rate) is called the Nyquist frequency of the sampling system.x[n] = x(nT). . the order is the larger of a in Eq. To handle this problem as gracefully as possible. The Nyquist±Shannon sampling theorem provides a sufficient (but not always necessary) condition under which perfect reconstruction is possible. It is possible under some circumstances to reconstruct the original signal completely and exactly (perfect reconstruction). most analog signals are filtered with an anti-aliasing filter (usually a low-pass filter with cutoff near the Nyquist frequency) before conversion to the sampled discrete representation.