Microphone

Microphone

Microphones fall into two main categories, Electret Condenser Microphones (ECM) and MEMS mcrophones. There're foil type (i.e. diaphragm type) and back electret type ECMs, and the critical difference is where the electrets are located. Foil type ECMs have electrets located on the diaphram and back electret type ECMs have electrets located on the back electret. Nowadays back electret types are more popular due to its superior performance. When it comes to MEMS microphones, we can see analogue MEMS mic (i.e. AMIC) and digital MEMS mic (i.e. DMIC). Customers are now accomodating themselves to DMICs. 

Regarding termination types, ECMs usually come with lead wires, possibly with connectors, or TH pins. On the other hand, MEMS mics are usually SMD. Since MEMS mic are situated on and thus limited by the PCBA location, different MEMS mics are designed to have different sound port positions to increase sound reception flexibility.  

According to directionality, microphones can be further divided into omni-directional, uni-directional and bi-directional mics. Omni-directional mics can receive sound coming at all angles, uni-directional with sound from the front, and bi-directional with sound in front and at back. In general, omni ones can suffice. It's because mics are often "inserted into" the housing, with its front exposed to the outside sound source and its back covered and sealed from outside. So actually due to the assemby method, most mics are "naturally" uni-directional. Only when uni-directionality cannot be achieved with the help of housing design, does one need to consider uni ones. Note that when you use uni mics, make sure the uni mic's back holes are exposed to the environment, in order to let uni-directionality take effect. On the other hand, there're no holes on the back of omni mics.

"Noise-cancelling" is a multi-faceted term and is open to different interpretation. But most agree that noise-cancelling can only be achieved by mic arrays. That is, two or more than two mics align with each other, and with the help of software circuit, they can cancel noise. In principle, by deploying arrays one can find out the noise and produce an anti-phase soundwave artificially to kill it. A mic component itself cannot achieve ideal noise-cancelling features. Some consider a mic as "noise-cancelling" because it's insensititve to low frequencies and high frequencies. As most noises are at extreme low or high frequencies, if a mic is insensitive to those bandwidths, then it can be loosely defined at "noise-cancelling". For example some bi-directional mics are deemed "nosie-cancelling". You can observe the mic's sensitivity by looking at its frequency response curve. For example if the mic is intended to receive human voice, of which the frequency is in hundreds to one or two thousand hz, then we want to see a curve with steep roll-off at both ends and a flat line in the middle bandwidth.

The capability of a mic to receive sound is defined as sensitivity, measured in dBV. Sensitivity is a negative number and the larger it is, the more sensitive a mic is. However, it doesn't mean we shall pursue a mic with the highest sensitivity. Note that if the sensitivity is too high, then you might inadvertently capture all the mild background noise. In a quiet environment you don't need a highly sensitive mic but in a noisy environment you might need it.

Apart from noise coming from the environment, there's also noise incurred by the mic's internal amp circuit. Here's a term called Signal-to-Noise Ratio (SNR). A high SNR means the electrical noise is small and vice versa. You'd always want a mic with higher SNR as it means the amp circuit is designed properly. Last, there's a term called Acoustic Overload Point (AOP). One may experience explosive noise coming from the mic because the sound source is way too loud. If AOP of a mic is high, it can process loud sound sources without creating explosive noise, for example, in a live concert or at an construction site.