Table of Contents
24bit ADC Selection Guide: Resolution, LSB, and Sampling Rate
Choosing an ADC is not only a question of "more bits." ADC resolution tells you the smallest ideal code step, while sampling rate tells you how often the converter can capture the signal. A 24bit ADC can be useful for low-level precision measurement, but it is not automatically better for fast control loops, audio, motor feedback, or general MCU sensing.
Start with four requirements: input range, smallest signal change you must see, signal bandwidth, and acceptable noise. Then choose the ADC bit depth, reference voltage, sampling rate, architecture, channel count, interface, package, and supply range that match the circuit. Inventory, price, lead time, certification, and lifecycle are dynamic sourcing fields, so they are supplier-dependent and must be checked against current distributor data.
Quick Selection Table
| ADC choice | Ideal codes | Best fit | Typical selection reason | Main caution |
|---|---|---|---|---|
| 8-bit ADC | 256 | Simple controls, coarse sensors, knobs, battery level bands | Low data size and simple MCU use | Large LSB; weak for small sensor changes |
| 10-bit ADC | 1,024 | General MCU analog inputs, simple sensors, embedded control | Better than 8-bit without much complexity | May still be too coarse for precision measurement |
| 12-bit ADC | 4,096 | Data acquisition, sensors, power monitoring, many external ADCs | Good balance of resolution, speed, and cost | Layout, reference, and noise begin to matter more |
| 16-bit ADC | 65,536 | Precision sensors, industrial measurement, medical front ends, weigh-scale subsystems | Smaller LSB and better measurement detail | Effective resolution depends strongly on noise and analog front end |
| 24-bit ADC | 16,777,216 | Low-speed high-resolution measurement, bridge sensors, precision instruments, delta-sigma ADC systems | Very small ideal code width for narrow-band signals | Do not expect 24 noise-free bits in a real circuit |
What Is ADC Resolution?
ADC resolution is the number of digital output codes available over the converter input range. An N-bit ADC has 2^N codes. For a unipolar ADC that maps 0 V to a full-scale reference range, the ideal code width, often called 1 LSB, is approximately full-scale input range / 2^N.
For example, a 12 bit ADC over a 3.3 V range has an ideal step size of about 0.806 mV. A 16 bit ADC over the same range has an ideal step size of about 50.4 uV. A 24bit ADC over the same range has an ideal step size of about 0.197 uV.
LSB, Reference Voltage, and Input Range
The reference voltage sets the scale for many ADCs. If the input range is 0 V to VREF, a smaller VREF creates a smaller ideal LSB, while a larger VREF supports a wider input range. Neither choice is free.
| Example input range | 8-bit LSB | 10-bit LSB | 12-bit LSB | 16-bit LSB | 24-bit LSB |
|---|---|---|---|---|---|
| 3.3 V full scale | 12.89 mV | 3.22 mV | 0.806 mV | 50.4 uV | 0.197 uV |
| 5.0 V full scale | 19.53 mV | 4.88 mV | 1.221 mV | 76.3 uV | 0.298 uV |
How Many ADC Bits Do You Need?
Use the smallest signal change you need to detect as the starting point. Required codes across range equal input range divided by the smallest useful change. Required bits are the base-2 logarithm of that code count.
If a 0 V to 3.3 V sensor output must resolve 2 mV changes, the ideal code count is 1650. That requires more than 10 bits because 10 bits gives 1024 codes, while 11 bits gives 2048 codes. In practice, a 12-bit ADC gives reasonable margin before noise, tolerance, and calibration are considered.
Sampling Rate Basics
Sampling rate is the number of conversions per second. It must be selected from the input signal bandwidth, not from the desired display update rate alone. The Nyquist rule says the sample rate must be more than twice the highest frequency component you want to represent.
That is only the mathematical lower bound. Real ADC systems need margin because anti-alias filters are not ideal, multiplexed inputs need acquisition time, digital filters add latency, and control systems often need timely updates.
Nyquist Is Not Enough: Use Real-World Margin
If the signal bandwidth is 10 kHz, Nyquist says sampling must be greater than 20 kSPS. That does not mean a 20.1 kSPS ADC is a safe choice. You still need filter transition band, timing margin, aperture jitter review for high-frequency signals, and enough samples for the algorithm.
8-bit vs 10-bit vs 12-bit vs 16-bit vs 24-bit ADCs
8-bit ADC
An 8 bit ADC gives 256 ideal codes. It can be enough for low-cost control and rough monitoring, but it is weak for small sensor changes.
10-bit ADC
A 10 bit ADC gives 1024 ideal codes and is useful for general embedded sensing when the analog front end is simple.
12-bit ADC
A 12 bit ADC gives 4096 ideal codes and is often a practical middle ground for sensors, power monitoring, and data acquisition.
16-bit ADC
A 16 bit adc converter provides 65,536 ideal codes, but the analog front end must support microvolt-to-tens-of-microvolts measurement quality.
24bit ADC
A 24bit ADC is often best for low-speed precision measurement where filtering, reference stability, layout, and calibration are controlled. It is not a universal upgrade.
Common ADC Selection Mistakes
- Choosing bits before input range.
- Treating Nyquist as the final sample rate.
- Ignoring reference voltage quality.
- Expecting 24 noise-free bits.
- Forgetting input drive and settling.
BOM Checklist for ADC Selection
| BOM field | What to specify | If missing |
|---|---|---|
| Resolution | 8-bit, 10-bit, 12-bit, 16-bit, 24-bit, plus required effective resolution | Examples: 12-bit MCP3208, 16-bit ADS1115, 16-bit AD7685 (SAR), 24-bit ADS1256 (delta-sigma) |
| Sample rate | Per-channel sample rate or output data rate | Examples: 860 SPS (ADS1115), 250 kSPS (AD7685), 30 kSPS (ADS1256), 200 kSPS (MCP3008) |
| Input range | Single-ended or differential voltage range | Examples: ±2.048 V (ADS1115 with PGA), 0-5 V (MCP3008), 0-5.25 V (ADS1256) |
| Reference | Internal, external, value, accuracy, drift, noise | Examples: internal 2.048 V (ADS1115), external 2.3-5.5 V (AD7685), external 2.5 V (ADS1256) |
| Interface | SPI, I2C, parallel, LVDS, JESD204, MCU-integrated, or other | Examples: I2C (ADS1115), SPI (MCP3008, AD7685, ADS1256), parallel (flash ADCs) |
| Procurement fields | Stock, price, lead time, lifecycle, certification | Must be confirmed with current supplier stock and lead time before BOM lock; all are active production parts as of August 2026 |
For project sourcing, browse ADC converter chips once the technical requirements are clear. For a multi-line project list, use Upload BOM with resolution, sample rate, input range, package, and interface notes included.
FAQ
What is ADC resolution?
ADC resolution is the number of output bits or codes available across the input range. An N-bit ADC has 2^N ideal codes.
How many ADC bits do I need?
Calculate the smallest useful signal change, divide the input range by that value, and convert the required code count to bits. Then add margin for noise, reference error, sensor tolerance, calibration, and effective resolution.
Is a 24bit ADC always better than a 16 bit adc converter?
No. A 24bit ADC is usually best for low-speed precision measurement. A 16-bit or 12-bit ADC can be better for faster signals, control loops, or simpler sensor systems.
What sampling rate should I choose for an ADC?
Start with more than twice the highest input signal bandwidth, then add real-world margin for anti-alias filters, multiplexing, acquisition time, digital filters, latency, and algorithms.
Sources
Alice lee
Business Manager
Focused on the electronic components sector, the author shares industry knowledge, product insights, and sourcing perspectives related to modern electronics manufacturing. With close attention to market trends, component applications, and supply chain developments, the content is designed to support engineers, buyers, and businesses in making more informed decisions.