The performance specifications provided for the IK Multimedia MEMS mic are somewhat rudimentary and no sensitivity is given.
The reference microphone was a calibrated PCB Piezotronics 376A32 0.5” phantom powered prepolarized condenser microphone (the subject of an upcoming review), which comprises a 377B02 capsule and a 426A14 preamplifier.
The other parts on the PCB are used to convert phantom power into an operational voltage suitable for the MEMS element and buffering circuitry to drive cables and a mic preamp (see Photo 3).Īll my measurements were done using an Audio Precision APx515 analyzer and an APx1701 transducer interface. The MEMS element includes the sensor itself, a buffer, and an output amplifier. The last point turns out to be moot, as we will soon see.Īlthough the MEMS element does not have a readable marking on it, it looks similar to the Knowles SPH1642HT5H-1, part of the SiSonic series, and referencing the datasheet for that part (see Resources) gives some insight into the operation of this family of elements (similar parts can be sourced from InvenSense and PUI, among others). Likewise, with this sort of mounting, the MEMS mic will not give accurate near-field results for woofer measurements. Photo 3: The circuitry is laid out on a double-sided PCB, with the MEMS element seen on the far right hand side (Photo courtesy of Cynthia Wenslow) The current MEMS mic is said to have an omnidirectional (free-field) polar pattern.
Running it through some basic tests, I quickly determined why IK Multimedia may have wanted to upgrade it-the top octave is severely rolled off (-10 dB at 20 kHz) and the distortion is relatively high. Ron was kind enough to send me the old mic, which appeared to be configured for diffuse-field measurement. The ARC 2.5 Room Correction System described by Ron Tipton in his review ( found here) makes use of a MEMS-based measurement mic, which replaces an electret-based mic used in earlier versions of the system. In the past few years, though, MEMS technology has improved to the point where it’s a viable option for test and measurement applications. This is likely the reason that electrets have still dominated in lower cost measurement application. The small diaphragm size has also resulted in a significantly higher noise floor.
Until recently, the disadvantages of MEMS have included low dynamic range and uneven frequency response due to the Helmholtz cavity resonance from the front and/or back cavity. Most MEMS work on the condenser principle, that is, they have a diaphragm and a back plate with a constant charge, and the acoustic pressures cause the distance between them (and hence the voltage) to be modulated. This not only makes the resulting parts far more consistent than usual mechanical assembly, but makes the integration of signal conditioning much simpler.
MEMS transducers generally are constructed in a similar manner as microchips (i.e., by using a piece of high purity silicon as a base and performing a series of etching and deposition operations). The small size has enabled their use in arrays, which is a great aid to DSP noise cancelling in many devices. They are also physically small and light and mechanically rugged. Compared with electret microphones, MEMS mics have a wide operating temperature range, good resistance to the stresses of soldering, and low current requirements. Microphones based on Micro-Electronic Mechanical Systems (MEMS) technology are now made in the billions, and totally dominate the huge cellphone, tablet, and pad market. Smaller, cheaper, better: time’s arrow for technology flies rapidly and inexorably. Photo 1: The new ARC 2.5 measurement microphone is built around an omni-directional, high precision MEMS capsule.