Which of the following are typical errors associated with ADC conversions?

• A. Quantization, resolution, linearity, offset, and noise
• B. Quantization only
• C. Offset only
• D. No errors

Answer: (A)

In an ADC, several error sources arise because a continuous signal is forced into discrete codes. The key ones you’ll encounter are quantization, resolution, linearity, offset, and noise. Quantization is the fundamental discretization error: the input is mapped to one of 2^N levels, so the value you read can be off by up to ±1/2 LSB, which shows up as quantization noise. Resolution, set by the number of bits, determines the size of each step; more bits mean smaller steps and finer distinctions, reducing that quantization error.

Linearity errors describe how the actual transfer curve deviates from a perfect straight line. Integral nonlinearity measures the overall deviation of the curve from the ideal line across the full range, while differential nonlinearity concerns the uniformity of successive step sizes. Offset is a constant shift in the output, so zero input might not map to the true zero code, which can bias all measurements if not corrected. Noise encompasses random fluctuations from various sources—thermal, flicker, amplifier noise, and others—and sets a limit on the smallest detectable signal. Quantization noise is a form of noise tied to the conversion process itself, but other noise sources also contribute to overall error.

This combination—quantization, resolution, linearity, offset, and noise—constitutes the typical error picture for ADC conversions. The other options miss some of these contributors, and saying there are no errors isn’t realistic for a real ADC.