Portable media players, like the Apple iPod and its competitors, have been around for many years, so you might think no further improvements are possible. But this doesn’t seem to be the case. Somehow, the companies that develop audio chips for these devices continue to come up with innovative ways to improve their offerings. The benefit to consumers is a more enjoyable listening experience for longer periods of time.
A slew of new audio chips promising high-quality audio and very low power consumption has surfaced since the beginning of the year, destined for a new generation of PMPs and cell phones. We highlight a selection of these chips here.
NEW CODEC ICS Introduced in February, the WM8900 from Wolfson Microelectronics (www.wolfsonmicro.com) is a high-performance, ultra-low-power audio codec that employs the company’s AudioPlus Smart Power technology (Fig. 1). The chip uses a class G, ground-referenced headphone driver.The WM8900’s quiescent headphone playback consumes less than 6 mW in voice mode and less than 11 mW in hi-fi mode, extending battery life in portable audio applications. Using a typical 300-mAh battery, the chip can push battery life out to 11 hours during headphone music playback at a typical 2-mW/ch listening level.
Wolfson also designed the WM8900 to address cost and board space through the aforementioned ground-referenced headphone outputs, which remove the need for bulky dc blocking capacitors. The device comes in a low-profile, 0.55-mm high, 40-pin quad flat no-lead (QFN) package—a good choice for slim, portable electronic applications.
The WM8900 class G amplifier architecture is implemented by powering the headphone amplifier with a dual-input, levelshifting, intelligent charge pump. This pump also generates both the positive and negative power-supply rails, ground-referencing the headphone outputs.
Automatic control of the charge pump maintains the most power-efficient operating state during headphone playback with no intervention required from the operator or host software. Also, the ground-referenced class G headphone amplifier eliminates many sources of pops and clicks during power-up, power-down, mute, and unmute to deliver high-quality audio performance and improved bass response.
“The WM8900 introduces class G amplifier technology as a means to deliver ultra-low-power, ground-referenced headphone drive on a hi-fi audio codec. The device helps designers of portable media players and multimedia handsets to meet the challenge of delivering longer battery life with reduced system cost,” said Nat Edington, vice president of marketing at Wolfson.
The codec uses stereo 24-bit, 64x oversampled sigma-delta analog- to-digital converters (ADCs). The multibit feedback and high oversampling rates reduce the effects of jitter and high-frequency noise. The ADC includes digital gain control, digital filtering, and a programmable digital high-pass filter. The ADC full-scale input level is proportional to the analog-supply voltage, AVDD. With a 3.3-V supply voltage, the full-scale level is 1.0 V rms.
Digital audio data is converted to oversampled bit streams in on-chip, true 24-bit digital interpolation filters. The bit-stream data enters two multibit, sigma-delta digital-to-analog converters (DACs) that convert the data to high-quality analog audio signals.
The multibit DAC architecture reduces high-frequency noise and sensitivity to clock jitter. It also uses Wolfson’s Dynamic Element Matching technique for high linearity and low distortion. Analog outputs from the DACs can then be mixed with other analog inputs using the WM8900’s output mixers. This mix is fed to the output drivers for headphone or line outputs.
The master clock can be input directly or generated internally by an integrated low-power, frequency-locked loop (FLL). The WM8900 operates at analog supply voltages down to 2.4 V. In addition, the digital core can operate at voltages down to 1.8 V to save power. Different sections of the chip can also be powered down under software control.
Continue on Page 2
Device performance includes a DAC to headphone signalto- noise ratio (SNR) of 95 dB (“A” weighted, 3.3 V) and DAC to headphone total harmonic distortion (THD) of –83 dB at 48 kHz Fs, 24 V. Also, the part has an ADC SNR of 92 dB (“A” weighted, 3.3 V) and ADC THD of –80 dB at 48 kHz Fs, 2.4 V. The WM8900 is available for sampling in a 40-pin, 5- by 5- by 0.55-mm QFN package.
Leadis Technology (www.leadis.com) also introduced a codec at January’s International Consumer Electronics Show—the LDS9350. This codec holds a “world’s first” title, since it’s the first portable audio codec with an integrated FM transmitter. Like the Wolfson codec, it uses class G amplifier technology to increase power efficiency and extend battery life in portable audio applications. The Leadis technology is called Gmax.
“Our line of Gmax-enabled codecs with integrated FM transmitters requires only 4 mW of power for quiescent playback,” says Greg Davis, senior director of Leadis’
Audio Business Unit. “We specify power consumption at average listening levels to provide a real-world estimate of battery life and to demonstrate the significant power-saving capability of the Gmax amplifier technology.” In addition to the stereo FM transmitter, the LDS9350 integrates a low-power and low-noise stereo audio codec, a highly efficient Gmax stereo headphone amplifier, a dc-dc converter, and a phase-locked loop (PLL) in a single compact solution (Fig. 2). Its 3.3-V, two-channel, 24-bit DAC has an SNR of up to 102 dB and can operate with sampling rates between 8 and 192 ksamples/s.
The FM transmitter function is fully programmable via an I2C interface with few external components. FM radio bands are selectable for worldwide operation: 87.5 to 108 MHz for the U.S. and Europe, and 76 to 90 MHz for Japan. Pre-emphasis and transmit power levels are addressable to meet country-specific regulatory standards. Audio to the transmitter can be digital via I2S or analog via stereo line-in or microphone inputs.
Leadis also offers codec-only versions, the LDS9302 (3.3 V) and LDS9302L (1.8 V single-supply), and a standalone FM transmitter, the LDS9200. The LDS9350, LDS9302, LDS9302L, and LDS9200 are available now. All are packaged in a standard 40-pin 6-by-6 QFN, as well as ultra-small chip-scale packaging (CSP). Single-unit pricing for 1000-piece purchases is $2.93 for the LDS9350, $2.38 for the LDS9302/L, and $1.95 for the LDS9200. An evaluation module includes a PC-based interface for easier programming and reduced time-to-market.
FRONT-END AUDIOAustriamicrosystems (www.austriamicrosystems.com) introduced a portable audio chip, the AS3543, at this year’s International CES, too. The company’s fourth-generation audio front end offers half the power consumption at twice the performance of the previous generation.
It’s designed to be an analog companion IC for advanced multimedia systemson- a-chip (SoCs) targeting media players, personal navigation, music phones, and generic mobile devices. The chip features a sub-7-mW stereo DAC at more than 100-dB SNR. It integrates audio and power management as well. Advanced powermanagement techniques like dynamic voltage with frequency scaling or adjustable dc-dc converters enable high efficiency over a wide range of current loads.
The AS3543 includes the stereo audio DACs and ADCs with line audio output and input, microphone input, headphone amplifier with direct drive capabilities, charging for lithium-ion and other popular types of batteries, efficient dc-dc step-up/ step-down power converters, low-dropout regulators (LDOs), ultra-low-power real-time clock (RTC), battery switch, and high-voltage backlight power unit.
Meanwhile, the AS3532 is the first member of a new family of media player ICs based on IP that bridges the gap between the audio experience of music phones and hi-fi home audio devices. The company touts the chip as an innovative musicplayer subsystem whose core is based on a newly developed audio engine and audio post-processor that act as co-processors to an ARM central programmable unit.
The audio engine, in a fully hardwired context, executes the decompression and playback of most popular compressed audio formats, like MP3, WMA, and AAC, for the least amount of power consumption with zero CPU load. The audio postprocessor implements an asynchronous sample rate converter (ASRC) with nearly transparent quality, a multichannel mixer with limiting function, and a 10-band graphic equalizer. It supports 192-kHz sample rates at a 24-bit dynamic range for high-definition audio processing.
Three sets of I2S outputs can independently control stereo speakers, subwoofer, and headphone or line outputs. They also can be used as multichannel audio outputs. On top of that, the audio subsystem includes a stereo pulse-duration-modulation (PDM) digital microphone input, thereby completing all audio requirements for new-generation mobile phones.
Buoyed by 512 kbytes of on-chip buffer memory, the AS3532 supports all key audio formats; compressed, lossless, and high-definition, polyphonic sequenced content; wavetable synthesizer; 3D positional sound; virtualizer engines; and other digital audio effects. There’s also a comprehensive software suite. The software development kit (SDK) has passed the stringent test criteria of the Certified for Windows Vista program for downloadable content.
The AS3532 is sampling to lead prospects in a 6- by 6-mm CTBGA package. Depending on the mobile-phone architecture, a matching high-definition integrated audio front-end and power-management component can be provided in the same 6- by 6-mm CTBGA package.