In ADC design, what is the effect of oversampling on quantization noise, and how does dithering help?

• A. Oversampling reduces noise density of quantization noise; dithering decorrelates quantization error, improving effective resolution.
• B. Oversampling increases quantization noise; dithering reduces thermal noise.
• C. Oversampling has no effect on quantization noise; dithering increases SNR.
• D. Oversampling converts quantization noise to clock jitter; dithering eliminates jitter.

Answer: (A)

Oversampling spreads quantization noise over a wider frequency range, so the portion of that noise that falls inside the band of interest is reduced. By sampling faster than the signal bandwidth and then filtering (decimating) down to the desired rate, the in-band quantization-noise density drops roughly in proportion to the oversampling factor, giving a higher effective resolution (an SNR gain of about 10·log10(OSR)).

Dithering adds a small, controlled random signal before quantization, which breaks the direct, input-dependent link to the quantization error. This decorrelation makes the quantization error behave more like white noise with a known variance, reducing distortion and tonal artifacts. When used with oversampling, dithering helps realize the benefits of noise being pushed out of band and processed by the digital filter, yielding a smoother, more linear effective performance and improved usable resolution.