Analog-to-Digital Converters (ADCs) vary in how long they take to complete a conversion, and whether that time depends on the specific magnitude of the signal being measured.
1. Successive Approximation Register (SAR) ADC:
The SAR ADC uses a binary search algorithm. For an $n$-bit converter, it always takes exactly $n$ clock cycles to find the digital value, regardless of whether the input is 0V, 5V, or any other value. Because it follows a fixed sequence (testing MSB down to LSB), the conversion time is constant.
2. Comparing Conversion Times:
• Counter Type: The counter starts at zero and ramps up until it hits the input voltage. If the input is low, it finishes quickly. If high, it takes long ($2^n$ cycles max). Time depends on input.
• Dual Slope: Integration time is fixed, but de-integration time depends directly on the input voltage magnitude.
• Flash ADC: While it is the fastest and finishes in 1 clock cycle (constant), the standard answer for "fixed conversion time" in balanced performance contexts is often the SAR ADC. However, Flash is also technically independent of the input voltage.
3. Why SAR is the primary answer:
In most standard electronics curricula, the SAR ADC is the classic example cited for having a fixed, predictable conversion time ($n \times T_{clk}$) that does not fluctuate with the input signal level.