Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
Aliasing in the context of signal processing is a direct inference from the Nyquist sampling theorem.
The Nyquist theorem states that any continuous time signal which is band-limited to a frequency Fc can be reproduced from its discrete samples if it is sampled at a rate Fs >= 2 * Fc.
Now, if this signal is is sampled at a lower rate than 2*Fc the continuous signal that is reproduced from the discretized samples will appear like a signal of a different frequency, in other words we can tell that the signal has suffered aliasing.
It is something like this when u sample a periodic signal ( a sine wave) then again try to build it using a Sample and Hold circuitry then u can get upto one particular level provided that there are sufficient samples to reconstruct the signal. The harmonics within the signal will still be present since any waveform is made up of a sequence of frequencies.
When u are sampling a signal below nyquist rate <2*max(freq) then in that cases the main signal is sampled with lot of subband frequencies as a result when u try to build the original waveform using this data then u might not get a clear waveform and u can see a distorted waveform this is concept of aliasing.
When you do a fourier plot for the waveforms u can see nearby frequencies overlapped or present in the basebnd signal this is concept of aliasing.
digital signal has a property that if any continous signal having freq Fm and it is sampled at sampling freq Fs, then digital signal contain interger multiple of Fs freq, K*Fs, K any +/- integer number.
That's why spectra of digital signal is multiple of Fs.
Now if Fs < 2*Fm, then spetra is overlapped at Fs/2 freq. Because of this overlapped spectra, information contained at near Fs/2 is loss. That means exact magnitude neat Fs/2 freq is loss. Lossing information is called alising.
To prevent alising, we have to pass signal through filter which has predifined response before sampling.
aliasing is when, higher frequency components takes the position or interrupts the lower frequency components
every sampling circuitry has a limit on what the input bandwidth is, (sampling rate/2 is the max input b/w, for ideal reconstruction). When the input bandwidth increases beyond this value, the reconstruction doesn't yield the input again, that is because the higher frequency components (which change rapidly, wrt sampling) have taken the position/Interrupted lower freq components.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
