Although we live in an analog world and possess organs that work in ways more analog than digital, it is common to find that we are converting most of the signals we handle to digital forms. This is because working with numbers is more convenient and we can have more uniformity in the digital circuits that process these signals.
However, the human ear does not take kindly to numbers, appreciating music only when it hears it in the analog. Fortunately, converting digital signals to analog is not difficult. Therefore, the entire chain of processing an audio signal can be in the digital domain, except at the extremities. These would be the sensors that collect the sound signals before processing and the transducers that convert the digits to analog sound after it has been processed.
Now, people are trying to convert to digital those entities at the very extremes of the audio processing chain. The humble microphone that has journeyed from the ribbon type to the electret type is now undergoing another makeover. Texas Instruments first introduced the digital microphone as a consumer device that could connect directly to TI’s codecs, bypassing the usual analog-to-digital converter in between. Where analog microphones picked up only 5KHz of the spectrum, digital microphones can now collect information over the entire 20Hz to 20KHz audio band.
That leads to many further possibilities. With digital microphones or an array of them, not only does audio sound much better than what we are used to, we can have sophisticated noise cancelling and beam forming for tracking a user’s voice.
Conventional electret microphones use a moveable diaphragm to form a capacitor at the gate of a JFET. Sound waves moving the diaphragm cause the capacitance value to change and the charge on it to vary. The JFET converts this varying charge to voltage that is amplified by operational amplifiers before being processed further.
That does not allow the electret microphone to get very much smaller and the signals from the JFET and the operational amplifier are prone to interference and noise. By using MEMS technology, not only can the physical size of the microphone be made more suitable to that demanded by the ever-shrinking digital devices, the ADC can be incorporated within the body of the microphone itself.
Therefore, we have a tiny digital microphone that produces a digital output either in a pulse density modulated (PDM) or in I2S format. The output can directly connect to a digital IC for further processing, eliminating all the issues related to interference and noise pickup.
Among the several types of semiconductor devices, the MEMS microphone package is unique in having a hole for the acoustic energy to travel to the transducer element. Within the package, the MEMS transducer and the ASIC are bonded together, being mounted on a common laminate.
The laminate has a lid over it to enclose the transducer and the ASIC. The laminate is actually a small PCB that routes signals from the electronics to the pins on the outside of the microphone package.
