In AD conversion, the analog signal is sampled for a period of time and then gets converted into a numerical value which CPUs can interpret. The resolution of the conversion is determined by the bit depth. This refers to how many possible digital values correspond to the range of the analog input signal. An 8-bit ADC can output 256 different levels, while a 12-bit ADC has 4096 output levels. An important parameter of AD conversion is the sampling rate, which determines how often the analog signal is converted.
SAR ADC is one of the most common and efficient AD conversion methods. Its operation is based on a binary search algorithm, meaning at a given bit depth, it steps the bits until it can create the most similar value to the measured value at that moment. It is commonly used in MCUs, embedded systems.
The Sample and Hold (S/H) circuit is responsible for taking a sample of the analog signal at a given moment and holding this value stable for a short period of time. During this time, a comparator quickly compares the sample with various references generated by the SAR circuit and DA converter. SAR is fast and low power.
In DA conversion, digital values are converted into an equivalent analog voltage or current. Perhaps the most common solution is the simulated PWM-based DAC, which converts a PWM signal into an analog signal using a filter. This is often used in embedded systems as it eliminates the need for separate DAC hardware.