Why is an anti-aliasing filter used before sampling in ADC design?

Why is an anti-aliasing filter used before sampling in ADC design?

• A. To amplify the signal to full-scale
• B. To limit the input signal bandwidth to below half the sampling rate, preventing aliasing
• C. To reduce quantization noise
• D. To convert voltage to current

Answer: (B)

Preventing aliasing when the signal is sampled is the main goal of an anti-aliasing filter. When you sample a continuous-time signal at a finite rate, the spectrum repeats every Fs, the sampling frequency. Any energy above half the sampling rate (Fs/2) can fold back into the baseband and distort the original signal after reconstruction. An anti-aliasing filter, a low-pass filter, attenuates frequencies above a chosen cutoff near or below Fs/2 so that out-of-band components don’t alias into the range you’re measuring. This keeps the sampled data representative of the true in-band content. The filter’s job isn’t to amplify the signal, and quantization noise is a property of the ADC itself, not something pre-filtering directly reduces. It also isn’t about converting voltage to current. By limiting the input bandwidth, you minimize spectral overlap and reconstruction errors, which is why this step is essential before sampling.

Preventing aliasing when the signal is sampled is the main goal of an anti-aliasing filter. When you sample a continuous-time signal at a finite rate, the spectrum repeats every Fs, the sampling frequency. Any energy above half the sampling rate (Fs/2) can fold back into the baseband and distort the original signal after reconstruction. An anti-aliasing filter, a low-pass filter, attenuates frequencies above a chosen cutoff near or below Fs/2 so that out-of-band components don’t alias into the range you’re measuring. This keeps the sampled data representative of the true in-band content.

The filter’s job isn’t to amplify the signal, and quantization noise is a property of the ADC itself, not something pre-filtering directly reduces. It also isn’t about converting voltage to current. By limiting the input bandwidth, you minimize spectral overlap and reconstruction errors, which is why this step is essential before sampling.