Audio Engineering Society

Chicago Section

Meeting Review, April 16, 2008


other meeting reports

4/16/08 Meeting Highlights
by Jeff Segota

The April 16th meeting of the Chicago Section featured a presentation on Advanced Condenser Microphone Simulation by guest Roger Grinnip, an Acoustical Engineer at Shure, Incorporated.  About 25 attendees were on hand to learn how advanced computational techniques can be applied to audio engineering.

 

A combination Boundary Element (BE)/ Analytical solver has been developed to simulate the response of both an omnidirectional and a unidirectional condenser microphone capsule.  This method is an improvement over traditional lumped element methods because the spatially dependent acoustical and mechanical components do not need to be approximated to one dimension.  

 

The calculation approach was described from a system perspective.  The capsule is comprised or four components having coupled responses.  The external surface pressure (with or without a port for omni/uni directionality) is modeled using the BE method.  The diaphragm displacement, diaphragm/backplate air-film pressure, and back chamber pressure are modeled using the analytical modal solution.  The BE mesh of the solid model is generated using commercial CAD software, and the response is simulated with custom C++ program. The frequency response to a free-field point source is calculated, and the spatial behavior of the diaphragm is animated. 

 

Comparing measured and simulated results of a prototype, initial attempts were good but there was deviation of a few dB in the frequency response above about 7 kHz.  And in the unidirectional case, the off-axis directionality was off.  Mr. Grinnip explained how he was able to improve the results when he considered the temperature profile inside the cavities, and this led to several questions.  Both simulated and measured (with a laser vibrometer) animations of the diaphragm displacement at 5 kHz were shown to have excellent agreement.

 

A healthy question and answer session concluded the meeting.  Topics included alternative methods for solving the matrix equation, simulating the self-noise, and expected results when various parameters are manipulated.