The architectural principle of the new generation audio DAC

This article describes the architecture of Wolfson Microelectronics' latest generation audio digital-to-analog converter (DAC), focusing on the design of new device families that provide high-voltage line driver output in consumer electronics applications.


Incremental accumulation modulators are often described in complex terms, using mathematical formulas, state tables, and theoretical models. Although all of this is necessary to understand the complexity of incremental accumulation modulation, the key to the purpose of this article is to understand the benefits of the SDM architecture and their application in audio converter ICs.

The two basic principles of incremental accumulation modulation are:

● Oversampling

The sampling process produces quantization errors; the difference between the sampling level at the output and the desired output level. The energy of the quantization noise depends on the resolution of the audio converter and is spread over the bandwidth of the sampling frequency.

The Nyquist sampling principle shows that in order to accurately convert a signal from the analog to the digital domain, the signal must be sampled at twice the frequency of the highest frequency component of the signal. The highest frequency component is also called the Nyquist frequency. For audio, the typical bandwidth is between 20Hz and 20KHz, and the sampling frequency tends to be 44.1KHz (for CD audio) to 192kHz (DVD audio).

A sampling frequency lower than twice the Nyquist frequency will cause aliasing, and the input signal will fold back to the audio frequency band with an image around the Nyquist frequency.

In SDM converters, data converters operate at frequencies much higher than twice the Nyquist frequency, usually 128 to 768 times the lowest sampling frequency.

The oversampling process diffuses the quantization noise over a wider bandwidth than other data conversion methods, so there is very little quantization noise in the audio frequency band.

● Noise shaping

In addition to spreading quantization noise over a wide spectrum, SDM is also used as a low-pass filter to filter the input signal. A high-pass filter filters the quantization noise, pushing the quantization noise out of the audio band. For ADCs, this allows the converter to use fewer bits without reducing SNR.

The requirement of oversampling means that the incremental accumulation modulator design is best suited for low bandwidth applications, such as audio data conversion, such as audio data conversion.

Design considerations

SDM-based architectures are complex, and designers have many options to optimize their designs for specific applications. The key tradeoffs are order, resolution, and architectural topology.

The order of the incremental accumulation modulator:

The first-order and second-order SDMs are inherently stable and generate large in-band noise, but have very low out-of-band noise. Higher-order SDMs can be conditionally stable and generate greater out-of-band noise, so they are sensitive to clock jitter.

The latest DAC architecture of Wolfson Microelectronics is based on a second-order incremental accumulation modem, which drives the clock at a high speed to reduce in-band noise and is therefore insensitive to clock jitter.

● DAC resolution

The increase in DAC resolution reduces the quantization error, thus improving the theoretical signal-to-noise ratio (SNR) of the DAC.

For each bit resolution, the theoretical maximum SNR is approximately 6xn, where n is the number of bits. Therefore, the theoretical maximum SNR of a 24-bit audio DAC is close to 144dB.

Wolfson ’s DAC design is based on a 5-bit or 6-bit converter, combined with the SDM architecture to provide a resolution of up to 24 bits.

For different noise sources, including analog and digital noise, the SNR cannot reach the theoretical maximum of -144dB. However, because of improved design methods, Wolfson's high-performance DAC of each generation strives to approach the theoretical maximum.

Performance, stability, size and cost are directly affected by the above design issues.

● DAC architecture

It can be considered that a typical incremental accumulation DAC contains the following elements: insertion filter—increasing the effective bit rate, allowing the DAC to oversample the input signal.

Wolfson uses a three-stage cascade integral comb filter (CIC) to attenuate the image from 8fs to 128fs. This method greatly attenuates frequency components that are several times the input sampling rate, improving the DAC's tolerance to clock jitter.

Incremental accumulation modulator-has the advantages of oversampling and noise shaping, which is critical for the high-performance audio data conversion described above.

Digital-to-analog converter—Converts SDM output to analog output. The switched capacitor method is used to precisely control the output voltage, and the noise introduced by the noise shaper is filtered to further improve immunity to clock jitter.

The patented method adopted by Wolfson includes a unique dynamic cell matching (DEM) scheme, which can minimize the capacitance mismatch error, compared with other alternative schemes, greatly improving the DAC linearity.

Low-pass filter—removes any remaining high-frequency components to achieve the most accurate reproduction of audio signals.

In fact, these four units are not completely isolated modules, and some functions are handled between these modules.

● Output level requirements

The audio DAC usually outputs a full-scale signal. Under the condition of 5V power supply, the level is between 1.0 Vrms and 1.1 Vrms. When the power supply voltage is 3.3 V, it is 0.66 Vrms and 0.72 Vrms. In mainstream applications, the output of the DAC is fed into an active circuit, which has two purposes:

Low-pass filter-it can remove the high frequency noise inherent in the conversion process.

Amplifier-The output level is usually increased to 2Vrms, which requires a high-voltage power rail (usually between 9V ~ 12V) to power the active devices of external circuits. There are several reasons for its implementation, including meeting industry standards, providing noise tolerance, and meeting de facto standards for interfacing with audio equipment.

Why is 2Vrms?

Various formal and de facto industry standards have been developed. These standards require the use of 2Vrms line levels between consumer audio equipment (such as DVD recorders) and televisions. However, there is a common misconception that the signal level must be 2Vrms plus or minus a specified tolerance. In fact, in most standard industry tests, audio equipment, such as DVD recorders, must be able to accept signal levels up to 2Vrms plus a tolerance. Devices such as DVD players must have an output level that cannot exceed 2Vrms plus tolerance. There is no minimum specification for the signal level, although to meet certain industry standards, the output signal level must be around 1Vrms.

In short, the audio transmitter must output a signal level between 1Vrms and 2Vrms, and the audio receiver must be able to receive input signals up to 2Vrms. Because the power supply of most consumer audio ICs is 5V or 3.3V, it is impossible to generate 2Vrms from a 5V power supply. Designers have to use external active devices to generate the required output level from the DAC circuit.

WM8501 and WM8522

Wolfson Microelectronics has considered these line level requirements in the design and designed a new series of audio devices. The output analog audio signals of these devices are 1.7Vrms line level and the power supply is 5V. This meets industry standards, eliminating the need for high-voltage power rails to boost the DAC output. This high voltage rail is still required in certain applications, such as SCART signal processing. But this is not required in all applications. Eliminating the 12V power rail and related power traces, active components of the DAC output filter and related PCB space, has realized a lot of cost savings and reduced circuit area.

Using a passive filter at the DAC output can get the same high-quality audio output. Compared with the active filter required to provide the necessary gain to amplify 1.0Vrms to 2.0Vrms, the complexity and cost are greatly reduced.

The WM8501 is a stereo audio DAC with a line driver output of 1.7Vrms. WM8522 is a 6-channel version of the same product, designed for multi-channel audio systems. Both chips can produce audio signals with a dynamic range greater than 100dB, which greatly exceeds any industry standard requirements, and therefore can meet the de facto standards set by consumer audio equipment critics.


Facts have shown that the design of incremental accumulation modulators is constantly evolving to meet market demands for performance and cost. In addition to improving the core architecture of their ICs, Wolfson Microelectronics is constantly discovering how to provide new ways for households to provide more benefits in performance, cost savings, and simplified terminal product design.

iPhone 4S Battery 3.7V Li-ion with 500 cycle life time which is make from brandnew cell and high quality protection board and FPC,all the raw materials is customerized. 0 cycle and iPhone 4S Battery life more than 500 times, rejection rate under 3‰. We are professional manufacturer on iPhone battery.

iPhone 4S Battery Pack

Nominal voltage: 3.7V

Limited charge voltage: 4.2V

Capacity:1430mAh (5.3whr)

Cell size: 42x32x82mm

iPhone 4S Battery

iPhone 4S Battery,iPhone battery,iPhone 4S Battery life,iPhone 4S Battery Pack

Shenzhen Aokal Technology Co., Ltd. ,

This entry was posted in on