•  
• EDN Europe MAGAZINE
•  
• Free SUBSCRIPTION
• Subscription Renewal
• View Media Kit

• EDN
• EDN Australia
• EDN Asia
• EDN China
• EDN Japan

Audio codec and headphone amp uses 2.5 mW

Dialog Semi’s class-G part includes on-chip digital filtering

Dialog Semiconductor’s DA7210 is a high-performance class-G headphone amplifier and audio codec that uses 2.5 mW quiescent, and 5 mW driving earphones at normal listening levels. The chip accepts a standard I2S 24-bit digital audio input stream at sample rates ranging from 8 kHz to 96 kHz, spanning all options from the most-compressed audio to hi-fidelity. It has stereo stereo playback/record paths, signal-to-noise ratio of 100dB DAC and 95dB ADC, and an integral I2C/PCM interface. It includes differential stereo microphone amplifiers. It is packaged in a 0.4mm-pitch QFN or CSP package with a 0.5mm-pitch CSP option (named the DA7211). Also on-board is a general purpose filter (GPF), a configurable DSP functional block that allows designers to adjust the output drive to, for example, compensate for the limitations of small, low-cost speakers and transducers. The DA7210 includes a true-ground, capacitor-less headphone driver that delivers 40mW output into 16 Ohm headphones.
Dialog says that customers in the portable-media-player market do not want class-D (switching) outputs stages, because of the difficulties of filtering the high-frequency spectral components from the output, but require more efficiency than is possible with class-AB amplfiers. Class-G uses a linear output stage but modulates the voltage rail that feeds the amp’s output stage; at low signal levels the voltage rail is reduced cutting quiescent current through the output drivers. Dialog says it has achieved added efficiency by, rather than modulating the output voltage rail linearly with signal level, setting it to a number of discrete voltage levels. Using multiples or sub-multiples of the basic voltage rail allows the DC/DC converter in the device to operate with maximum efficiency. Although the output-stage voltage rail varies in steps, the audio output contains no trace of switching effects, Dialog says. The chip’s circuitry must therefore “look ahead”, setting the output rail according to the signal levels in the audio stream. Dialog’s engineers say that the small latency of the signal path throught he codec provides sufficient time for their design to modulate the voltage ahead of the arrival of signal peaks.
The device operates down to 1.8V simplifying digital processor interfacing. The GPF enables onboard signal tailoring and performance tuning without degradation or loss of sound quality, and removes the need for a dedicated external DSP. Traditional low power audio codecs rely on external DSP devices running proprietary algorithms for audio enhancement or more complex noise suppression functions. Dialog's integral GPF comprises eight programmable bi-quad sections within the codec device itself. The audio signal sent to the speakers can be tailored for optimum response without modifying the host processor software. Engineering samples will be available from November 2009 with volume production delivery from Q1 of 2010; pricing will be $1.65 (10,000).