This article is about classic Q&A on ADC/DAC design, covering common issues such as clock duty cycle, common mode voltage, gain error, differential phase error, intermodulation distortion, and more.
1. What is Small Signal Bandwidth (SSBW)?Small Signal Bandwidth (SSBW) refers to the frequency at which the output amplitude drops to a specified value compared to the low-frequency output amplitude, given a specified input signal amplitude and frequency.2. What is Common Mode Voltage (VCM)?Common Mode Voltage (Common Mode Voltage (VCM)) is the same DC input voltage present on both pins of a differential input.3. What is MSB (Most Significant Bit)?MSB (Most Significant Bit) is the bit with the highest value or weight. Its value is half of the full scale.4. What is Sampling (Aperture) Delay?Sampling (Aperture) Delay is the time required from the trailing edge of the clock input to the opening of the sampling switch. The sample-and-hold circuit effectively stops capturing the input signal and enters the “hold” mode, determining the sample after the clock delay.5. What is Full Scale (FS) Input Range?Full Scale Input Range refers to the input voltage range that is digitized by the analog-to-digital converter, which should not be below or exceed this range. For example, V REF + = 3.5V and VREF – = 1.5V, FS = (VREF +) – (VREF -) = 2.0V.6. What is Clock Duty Cycle?Clock Duty Cycle is the ratio of the high-level time of the clock waveform to the total time of one clock cycle.7. What is Effective Number of Bits (ENOB)?Effective Number of Bits (ENOB) is the ratio of signal-to-noise ratio and distortion, or another expression of SINAD. ENOB is defined as (SINAD – 1.76) / 6.02, and this number (ENOB) indicates how the converter is equivalent to an ideal analog-to-digital converter.8. What is Gain Error?Gain Error is the difference between the actual input voltage and the ideal input voltage when the first and last codes are converted. This difference is: full scale – 2 LSB.9. Many analog-to-digital converters provide applications in the datasheet, with three capacitors appearing on the Va, Vd, and Vref pins. Are all three capacitors necessary?Depending on the specific device and the layout of the circuit board, one or two capacitors may be sufficient. Larger capacitors, typically 5 to 10 µF, provide low impedance bulk storage to ensure voltage stability during conversion. Smaller capacitors absorb high-frequency noise spikes. If the printed circuit board has a very good layout for low-noise operation and does not contain a microcontroller or other noisy digital logic, fewer capacitors may be needed. However, to ensure the ADC operates with level accuracy, it is generally best to follow the recommendations in the datasheet.10. What is Zero Error?The zero error of the ADC’s dual output is the difference between the theoretical input voltage (typically the midpoint plus 1/2 LSB) and the actual input voltage that causes the output to transition from 0 to 1.11. What is Output Hold Time?Output Hold Time refers to the duration of valid data output after the input clock edge.12. What is Resolution?Resolution is the analog increment equivalent to a change of 1 LSB in the converter code. Resolution is also defined as the number of bits (n) of the converter. The number of digital codes equals 2^n, where “n” is the number of bits. For example, a 12-bit converter maps an analog signal to 2^12 = 4096 digital codes. The resolution of a 12-bit analog-to-digital converter is the full-scale input voltage divided by 2^12, or 4096, ensuring that the output code does not exceed the range.13. What is Differential Phase Error?Differential Phase Error refers to the difference in output phase when reconstructing a small signal sine wave at two different DC input levels.14. What are the Dynamic Specifications of an Analog-to-Digital Converter?The Dynamic Specifications of an analog-to-digital converter involve those AC input signals. These include Signal-to-Noise Ratio (SNR), SINAD (Signal Noise + Distortion), ENOB (Effective Number of Bits), THD (Total Harmonic Distortion), IMD (Intermodulation Distortion), FPBW (Full Power Bandwidth), and SSBW (Small Signal Bandwidth).15. What is Intermodulation Distortion (IMD)?Intermodulation Distortion (IMD) refers to additional spectral components that do not appear at the input when two sine waves of different frequencies simultaneously act on the input of the analog-to-digital converter. It is defined as the ratio of the energy in the intermodulation products to the total energy in the original frequencies. Intermodulation Distortion (IMD) is typically expressed in decibels (dB).16. What is Gain Temperature Coefficient (Full Scale Temperature Coefficient)?Gain Temperature Coefficient (Full Scale Temperature Coefficient) refers to the ratio of the change in gain error to the change in temperature. It is usually expressed in parts per million per degree Celsius (ppm/°C).17. What is Total Harmonic Distortion (THD)?Total Harmonic Distortion (THD), expressed in dB or dBc, refers to the total harmonic level (the National Semiconductor analog-to-digital converter has 9 harmonic bands) and the level of the input signal’s harmonics appearing at the output. The calculation method for Total Harmonic Distortion (THD) is as follows: THD = “sqrt”[ (f2xf2 + f3xf3 + f4xf4 + f5xf5 + f6xf6 + f7xf7 + f8xf8 + f9xf9 + f10xf10) / (f1xf1) ], where f1 is the fundamental frequency of the input signal, and f2 to f10 are the harmonic frequencies that are multiples of the fundamental frequency.18. What is Zero Scale Offset Error?The zero scale offset error of a unipolar output analog-to-digital converter is the difference between the ideal input voltage (1/2 LSB) and the actual input voltage that causes the output code to transition from 0 to output code 1.19. What is Full Power Bandwidth (FPBW)?Full Power Bandwidth (FPBW) refers to the frequency at which the full-scale input drops to 3 dB below its low-frequency value in the reconstructed output.20. What is LSB (Least Significant Bit)?LSB (Least Significant Bit) refers to the smallest value or weight among all bits. This value is m*VREF/2^n, where “m” is the main reference range factor (this is the most common unit), and “n” is the resolution of the analog-to-digital converter.21. How can overshoot and/or undershoot on the clock and control lines of the analog-to-digital converter be eliminated?Overshoot and/or undershoot are caused by high-speed signal edges and mismatched signal terminations. Adding a 47-100Ω resistor in series to the input, as close to the clock source as possible, is recommended. The goal is to match the signal impedance on the clock line, and transmission lines should also be considered in this regard. A small additional power supply for the series termination is usually sufficient to reduce ringing effects.22. What is Output Delay?Output Delay refers to the delay time from the trailing edge of the input clock to the appearance of data updates at the output pin.23. What is Aperture Jitter?Aperture Jitter refers to the variation in aperture delay between sampling values. Aperture jitter appears as input noise.24. What is Signal-to-Noise and Distortion Ratio (S/(N+D) or SINAD)?Signal-to-Noise and Distortion Ratio (S/(N+D) or SINAD) is expressed in decibels (dB) and is the ratio of the effective value of the input signal at the output to the effective value of all other spectral components, including harmonics below half the clock frequency, but excluding DC.25. What is Full Scale Step Response?Full Scale Step Response is defined as VIN changing from VREF- to VREF+, or from VREF+ to VREF-, allowing sufficient recovery time for the converter to meet its rated accuracy for the conversion.26. What is Channel Delay (Latency)?Channel Delay (Latency) refers to the number of clock cycles between the start of conversion and when the conversion data appears at the output driver stage. Data for any given sample is available after the channel delay plus output delay. New data can be obtained every clock cycle, but the data lags behind the channel delay plus output delay.27. What are the Static Specifications of an Analog-to-Digital Converter?Static Specifications are metrics related to the DC (Direct Current) signal inputs of the analog-to-digital converter. These include gain error, offset error, and differential and integral linearity errors.28. What is Total Unadjusted Error (TUE)?Total Unadjusted Error (Total Unadjusted Error (TUE)) refers to the maximum deviation of the digital code center from the input voltage range under ideal conditions. Total Unadjusted Error includes offset error, gain error, and differential and integral non-linearity errors.29. What is Full Scale Error?Full Scale Error is the measurement of how far the last code conversion is from the ideal value of 1/2 LSB below VREF +, defined as: VFSE = Vmax + 1.5 LSB – VREF +, where Vmax is the voltage converted to the maximum code, which can be expressed in volts, as a percentage of the least significant bit or full-scale range.30. What is Conversion Time?Conversion Time refers to the time required for the analog-to-digital converter to complete one conversion. Conversion time does not include sampling time, multiplexer setup time, or other parts of completing a conversion cycle; conversion time may be less than throughput time.31. What is Power Supply Rejection Ratio (PSRR)?Power Supply Rejection Ratio (PSRR) can be divided into two specifications. DC PSRR is the ratio of the change in a specific parameter (e.g., full scale error) to a specified change in power supply voltage. AC PSRR measures the amplitude of a specific frequency and amplitude signal superimposed on a power supply, comparing its output amplitude at the output to its amplitude at the power supply pin. Power Supply Rejection Ratio (PSRR) is typically expressed in decibels.32. What are Missing Codes?Missing Codes are those output codes that are ignored or will never appear at the output of the analog-to-digital converter. These codes cannot be achieved by any input value.33. What is Throughput Rate?Throughput Rate is the maximum continuous conversion rate of the analog-to-digital converter.34. What is Signal-to-Noise Ratio (SNR)?Signal-to-Noise Ratio (SNR) is a ratio expressed in decibels (dB) that represents the effective value of the input signal at the output compared to the effective value of all other spectral components (below half the sampling frequency, excluding harmonic components and DC). SNR is the decibel number of the effective value of the signal level compared to the effective value of various noise (including quantization noise, thermal noise, white noise, etc.). The signal refers to the effective value of the fundamental component, while noise refers to the effective value of all non-fundamental components below the Nyquist frequency (excluding harmonic components and DC).35. What is Throughput Time?Throughput Time refers to the time taken by the converter to complete one conversion. Throughput Time includes any multiplexer setup time, sampling time, conversion time, and output display time.36. What is DC Common-Mode Error?DC Common-Mode Error is a specification for the differential input of the analog-to-digital converter. It refers to the change in output code when the same value is added to the analog voltages on both inputs.37. What is Offset Error?The offset error of the ADC is defined as the difference between the input voltage that causes the least significant bit to be set to “1” and the theoretical input voltage that would cause the least significant bit to be set to “1”.38. Should the power ground of the printed circuit board be a common ground for both digital and analog?Yes, they should be the same. However, it is important to carefully route all power and signal wiring to keep ground currents separate from power and signal.