Today’s consumer electronics devices are on the cusp of an audio revolution. While in recent years designers have focused on the development of exciting new functions such as wireless Internet access and mobile TV, advances in audio functions have lagged behind. That trend is about to change.

A variety of enhancements in microphone technology, including higher signal-to-noise ratio (SNR), flatter wideband frequency response, and improved sensitivity and phase matching promises to drive the development of new audio features from HD audio and wideband Voice-over-IP (VoIP) to improved audio/video recording and beam forming for hands-free communication.

Driving this trend is a growing recognition by system designers that poor microphone performance at the front of the audio processing signal chain has a far-ranging impact on overall audio quality. Audio subsystems built around a microphone with limited performance significantly increase the requirements for downstream processing to condition and improve the audio signals.

These requirements, in turn, translate into higher power consumption, higher system and development cost, and a more complex system design. Furthermore, despite the best of efforts, audio quality will be limited by the performance of the microphone used for audio capture. Inevitably, poor microphone performance limits designers’ ability to enhance their end product performance with new audio capabilities such as beam forming, noise cancellation techniques, or stereo sound.

Performance Limitations

Most audio subsystems in today’s consumer and communications devices use electret condenser microphones (ECMs). These devices comprise a fixed backplate with a non-conductive precharged layer attached and a flexible membrane, usually made of a metalized Mylar. The backplate and the membrane, which moves in response to sound, form a capacitor.

Movement of the membrane changes the capacitance, resulting in a change in output voltage. A small three-terminal JFET mounted inside the microphone canister serves as a buffer between the capacitive sensor and the output. An external pre-amplifier is often used to deliver the signal to an analog-to-digital converter (ADC).

Historically, system designers have employed ECMs because they are readily available from multiple suppliers and have offered an affordable solution. Most recent ECM development efforts have focused on lowering cost and reducing footprint, but manufacturers have achieved little success improving microphone sensitivity, SNR, and linearity.

As a result, while ECM technology provided an adequate solution for consumer electronics applications in the past, today it presents a number of performance limitations. The technology consumes relatively high amounts of power, a key concern for designers of battery-powered mobile systems. Power-supply rejection (PSR), or the ability to suppress noise on the power supply, is also relatively poor.

The poor PSR limits a designer’s flexibility in placing the microphone due to concerns about noise generated from other system components such as the LCD. Designers using ECMs can compensate for the technology’s poor PSR by adding a low-dropout regulator (LDO) to generate clean power for the microphone. Yet this approach increases system component count, driving up system footprint, power requirements, and cost.

Moreover, ECM technology entails some additional hidden costs. First, the use of electrets often requires hand assembly, which adds additional time and cost to the manufacturing process. ECMs also require a number of other supporting components such as a discrete converter and a preamp. These additional devices drive up board real estate requirements and increase power consumption and cost.

Furthermore, electrets cannot offer the tight tolerances and repeatable part-to-part performance that system designers enjoy with components manufactured in today’s photolithography-defined semiconductor processes. The sensitivity and frequency response of an ECM can vary significantly from device to device as well as over temperature, making it difficult for system designers to match components for basic applications such as stereo.

To compensate for this weakness, manufacturers building multi-microphone designs for such applications must often hand-sort ECMs in an attempt to better match components. This, in turn, drives up costs and further complicates the manufacturing process.

System manufacturers have begun shifting to surface-mountable electrets to eliminate some of these issues. However, the industry-wide migration to lead-free soldering has presented additional obstacles. Higher reflow temperatures used in lead-free manufacturing degrade electret performance, and manufacturers often have been forced to run multiple reflow operations to reduce the impact of the reflow process on electret performance.