# How A/D Converter Works? ICs Explained

Follow article### Introduction

Analogue to Digital Converter (ADC) is a very important component in electronic equipment. Since most real-world signals are analogue, ADC converting interfaces are necessary to allow digital electronic equipment to process the analogue signals. It converts any analogue signal into quantifiable data, which makes it easier to process and store, as well as more accurate and reliable by minimizing errors.

### 1 Understanding Analogue-to-Digital Conversion

The analogue-to-digital converter is mainly to sample the analogue signal, and then quantize and encode it into a binary digital output. The following are specific converting these steps:

*1) Sampling and Hold Circuit*

Due to the short sampling time, the sampling outputs a series of intermittent narrow pulses. It takes a certain amount of time to digitize each sampled narrow pulse signal. Therefore, between two samplings, the sampled analogue signal should be temporarily stored until the next sampling pulse arrives, this action is called hold.

In the design of analogue circuits, it is necessary to add a sampling hold circuit, To ensure correct conversion, the analogue circuit must retain the data that has not yet been converted. A sampling hold circuit can ensure stable sampling time in the analogue circuit. In addition, capacitor components are usually used to store charge.

*2) Quantizing and Encoding*

The quantizing and encoding circuit is the core part of the ADC. Generally, there are two ways to quantize the sampled value:

a. Quantify

First, take a minimum quantization unit Δ=U/2n (U is the maximum value of the input analogue voltage, and n is the number of bits of the output digital value). When the input analogue voltage U is between 0 and Δ, it is classified as 0Δ, and when U is between Δ and 2Δ, it is classified as 1Δ. The maximum quantization error produced by this method is Δ/2, and the error is always positive +1/2LSB.

b. Quantify and Carry Save

If the quantization unit Δ=2U/(2 n+1–1) when the input voltage U is between 0 and Δ/2, it is classified as 0Δ, and when U is between Δ/2 and 3/2Δ, it is classified as 1Δ. The maximum quantization error produced by this method is Δ/2, and the error is positive or negative, which is ±1/2LSB. So the quantization result also causes the so-called quantization error.

*3) Resolution*

Refers to the smallest analogue input that the A/D converter can distinguish. It is usually expressed by the number of bits converted into a digital quantity, such as 8-bit, 10-bit, 12-bit and 16-bit. The higher the number of bits, the higher the resolution. Any change in the input analogue voltage that is less than the minimum charge will not cause a change in the output digital value. However, it is very important to choose an applicable A/D converter, not that the higher the resolution, the better. Where high resolution is not required, most of what is captured is noise. In addition, the resolution is too low, the desired signal may not be sampled.

*4) Conversion Error*

This is usually the output in the form of relative error, which represents the difference between the actual output digital value of the A/D converter and the ideal output digital value and it is expressed in multiples of the least significant bit LSB.

*5) Conversion Time*

The conversion time is the time required for the A/D conversion once a time. In other words, the time interval from the start to the end of the conversion and a stable digital output value is obtained. The shorter the conversion time, the faster the conversion speed.

*6) Accuracy*

For A/D converters, accuracy refers to the difference between the actual required analogue input value and the theoretically required analogue input value when the set digital value is generated at the output. According to different calculation methods, the accuracy can be divided into absolute accuracy and relative accuracy.

### 2 Which ADC Architecture is Better?

The structure usually adopted by ADC includes the following 4 types:

1) Parallel structure, including Flash ADC

2) Segmented structure, including folding and interpolating ADC

3) Iterative structure, including partition ADC, pipelined ADC, and successive-approximation ADC

4) Sigma-delta (Σ-Δ) structure, including Σ-△ ADC

### 3 What is an ADC Input?

1) Analogue input, which can be single-channel or multi-channel analogue input.

2) Reference input voltage, which can be provided externally or generated inside ADC.

3) Frequency input, usually provided by the outside, used to determine the conversion rate of the ADC.

4) Power input, usually have analogue and digital power interfaces.

5) Digital output, ADC can provide a parallel or serial digital output.

### 4 What is an ADC Chip?

Due to the different working principles and process technologies used to achieve analogue-to-digital conversion, a wide variety of ADC chips are produced.

According to the resolution, the A/D converter is divided into 4 bits, 6 bits, 8 bits, 10 bits, 14 bits, 16 bits and 31/2 bits and 51/2 bits of the BCD code. According to the conversion speed, it can be divided into super high speed (conversion time=330ns), sub-super high speed (330~3.3μS), high speed (3.3~333μS), low speed (>330μS) and so on.

ADC can be divided into the direct type and indirect type according to the conversion principle. The former directly converts analogue signals into digital signals, such as successive-approximation ADC, parallel ADC and so on. Among them, the successive-approximation ADC is easy to implement with integrated technology and can achieve higher resolution and speed. Indirect A/D converter converts analogue quantity into intermediate quantity, and then into digital quantity, such as voltage/time conversion type (integral type), voltage/frequency conversion type, voltage/pulse width conversion type, etc. Among them, the integral A/D converter has a simple circuit, strong anti-interference ability, and can achieve high resolution, however, its conversion speed is slow. Some converters also integrate multiplexers, reference voltage sources, clock circuits, decoders and conversion circuits into one chip, which is very convenient to use.

ADCs are often used in communications, digital cameras, and computer systems to facilitate digital signal processing and information storage. In most cases, the ADC function will be integrated with the digital circuit on the same chip, but some devices still need to use an independent ADC. For example, mobile phones integrate ADC functions in digital chips, and cellular base stations with higher requirements need to rely on independent ADCs to provide the best performance.

Here lists several ADC ICs and their features and specifications as ADC selection references:

1) AD7621: 16-Bit, 2 LSB INL, 3 MSPS PulSAR® ADC, High sampling rate, Available in a 48-lead LQFP or a 48-lead LFCSP (708-9971)

2) AD7641: 18-Bit, 2 MSPS, Charge Redistribution SAR ADC (708-9971)

3) AD7908: 8-Channel, 1 MSPS, 8-Bit ADC with Sequencer in 20-Lead TSSOP (183-1936)

4) AD7918: 8-Channel, 1 MSPS, 10-Bit ADC with Sequencer in 20-Lead TSSOP (538-6517)

5) AD7928: 8-Channel, 1 MSPS, 12-Bit ADC with Sequencer in 20-Lead TSSOP (538-6523)

6) AD5555: Precision DUAL 16-Bit 14-Bit-DACs in Compact TSSOP Packages

7) AD8230: 16V Rail-to-Rail, Zero-Drift, Precision Instrumentation Amplifier (709-5070)

8) AD7799: 3-Channel, Low Noise, Low Power, 24-Bit, Sigma Delta ADC with On-Chip In-Amp (041-2315)

## Comments

Add a comment