To speed iterative prototyping and fixture creation for some assemblies and housings, IAR has brought reliable and low-cost FDM prototyping in-house. Common items we print include speaker baffle adapters, customer housing mockups for probe microphone measurements, or other fixtures to support device assembly or measurements. Maximum print size on this device is approximately 9 x 8 x 8 inches. High-resolution SLA and smaller assemblies will continue to be sourced from a handful of Indianapolis-based prototyping shops.

 Many far-field voice devices employ beamforming microphone arrays to improve speech recognition and communication performance. However, often the simulation of the acoustic wave during DSP design only includes phase differences due to array spacing and neglects more complex geometry such as element porting, enclosures, tables or walls.  These objects cause diffraction and reflections of the incoming acoustic wave around the sensor can lead to errors in beamforming and direction-of-arrival algorithms. IAR can use Comsol Multiphysics or Lumped Element Simulation (depending upon the complexity of geometry, frequency range of interest and available time) to provide simulated complex acoustic pressure “vectors” to DSP designers to improve the performance of the array including geometric features, prior to any prototype production or PCB fabrication.

Many wireless phones tout their water resistance these days.  BUT, for microphones, there are several potential problems the designer must be aware of before using environmental shielding barriers.  These barriers or membranes that give so-called “better than IP67 protection” (see the IEC 60529 dust and water ingress protection standard) have both acoustic mass and acoustic resistance which will affect the microphone performance.  The main effect for omnidirectional mic installation is that the sensitivity will be lowered a small amount and the upper end of the response can be changed due to shifts in acoustic resonances.  Much more drastic effects can be seen in arrays of omnidirectional or gradient/differential mics.  In these mics, the mass and resistance and the associated variation of these properties can severely shift the polar/directional characteristics and the sensitivity.  These shifts in performance are due to the subtraction of the pressures electrically in the case of omni arrays or acoustical subtraction in the case of a gradient mic.  Subtracting large numbers to get small numbers as in these cases yields small numbers that still carry the large variation effects.  So, besides matching omnidirectionals for sensitivity and response for arrays (not even mentioning the other confounding issue, self-noise) or selecting a good gradient/directional mic, careful attention must be paid to the surrounding acoustical elements, INCLUDING THE BARRIER ELEMENTS, in any design.

By: Marc Reese and Larry Marcus
​
Indy Acoustic Research celebrates its first year! 

A lot of effort has gone this year into building our lab's capabilities to ensure we can support our customer's needs and timing. Below are some pictures of some of our activities in the past few months.

by: Larry Marcus

Built in 1984, the IAR Anechoic Chamber is like me: it’s old but it still works. It has a cutoff frequency of about 120Hz and is about 3 meters by 3.7 meters by 2.6 meters tip-to-tip.  Compare the photo here with the more recent one below in this blog series with the four intrepid IAR founders!

Significant products developed and studies conducted in our anechoic chamber include many IEEE and TIA standards contributions, the first gradient microphones with speakerphones, first extensive ISDN and VoIP testing, first mechanical-acoustical analysis of conference phones, etc., not to mention days upon days of other transducer development and product testing.

​Excel is a great tool for engineering data reporting due to its widespread use. However, it has never supported a good polar plot which is an essential plot type in multiple engineering disciplines, including acoustics.  The major workarounds I’ve seen are to either download a plugin (which is difficult if you don’t have admin rights) or convert (theta,R) into x=R*cos(theta) and y=R*sin(theta) and use a scatter plot. However, the latter method can take up a lot of spreadsheet space -especially if you have multiple sets of data.
 
It turns out the chart type that you always thought should be able to plot a polar – the Radar Chart – CAN do it; however, there is still some frustration with the formatting of axes. Thankfully, if you work through the frustration once and save the completed chart as a template, it’s much easier to reuse future data sets. See the below attachment for instructions using Excel 2016. There are some differences with older versions, but hopefully you can follow along also...
​

IAR has moved back into the old lab!  The lab has a fully anechoic chamber for far-field devices, Head and Torso simulator (HATS) for near field devices and diffuse field booth for noise isolation measurements. And yes, some desk space for sippin' coffee.