Several design challenges exist when designing audio for a telecom handset. These include:
For a telecom handset to be type approved, certain performance criteria must be met. The handset's frequency response (receive) must be measured on a standardized artificial ear, which basically is an acoustical copy of the human ear designed with a controlled leakage. The frequency response must lie within a certain "window" defined in the standard. The same goes for the outgoing sound (send) which also has to meet certain frequency response criteria defined in the standards. Examples of such windows are shown below.
The share of handsets that comply with the Super Wideband standard is growing, which requires receivers with high technical performance as well as it increases the design complexity of the handsets.
When holding a handset close to the ear there will always be an acoustic leakage formed between handset and the ear. If the receiver was ideal (large sized membrane & motor), such leakage would not degrade the sound quality. In most modern handsets the performance of the incoming audio has shifted from being handled by the transducer alone to involve acoustical access to the internal part of the handset. This is due to the fact that handsets are getting smaller and smaller which calls for smaller components, including the receiver (driver) producing the sound. However, shrinking the receivers size will have a huge impact on the sound reproduction capability easpecially in the bass region if no modifications are done.
Example of frequency response of handset without leak tolerance when held firmly to the ear (solid) and when moved just a few milimeters away from the ear (dashed). A leak tolerant design would only loose a few dB's and thus be similar to the solid curve in both situations resulting in a much more consistent (pleasant) LF reproduction.
The solution is to use a receiver designed to have a huge bass boost and introduce a controlled leakage between the front of the receiver to the outer free field or alternatively to the inside of the handset (rear side of the driver). The result will be that the huge initial bass boost will be loaded down to a flat response, which then can accept an additional leakage between handset and ear without any significant loss of bass content. This design however is not not straght forward and involves controlling acoustic resistances, masses and volumes using slits, tubes, cavities and fabric (damping).
The leakage path can be built into the handset using carefully dimensioned tubes and acoustic fabric.
The OWR-3083 receiver has built-in leakage which can simplify the acoustic design of the handset.
In most cases the receiver unit will produce this magnetic signal using its built-in voice coil. Alternatively an external tele coil can be built into the handset if the receiver cannot generate the required magnetic signal strength on its own. Receivers having HAC compatibility are typically designed to have a “high" current coil for keeping up a high magnetic signal as well as a high acoustic signal.
It is important that the receiver's frequency response (in the HAC window) is acoustically well balanced since drastic DSP corrections would also affect the HAC signal.
Typically space is not limited in the bottom area of the handset where the microphone is positioned. The microphones are typically electrete microphones in the size range of Ø6mm to Ø10mm which are low cost "work horses" with great performance. If space is limited, a much smaller MEMS microphone is the ideal choice.
For both microphone types the HF response can be tuned by the front cavity and sound inlet in the handset, forming an acoustic low pass filter. Sometimes a front resonance at, say, 10 kHz can be designed into the handset to boost the response in that frequency range. The microphone response (send) must also comply with standards as described in the section "Target response and standards" above.
Read more about Ole Wolff microphones
Using Comsol Multiphysics™ Ole Wolff can simulate all of the acoustics involved when designing a telecom handset. We use accurate models of artificial ears for comparing simulation results with experimental data. Typically we do "virtual prototypes" in close cooperation with the customer's design team. In this way the tranducers, acoustic chambers etc. can be modified and changed without ever having to build physical prototypes. Once the simulation results show that performance targets are met, samples are built and tested in the lab.
Speakers for hands-free operation also need to comply with telecom standards. Examples shown in the chart below.
Narrow band (dashed) and Super Wide Band (solid) frequency windows used for hands-free audio
The OWR-3083 is a HAC compliant receiver with built-in leakage which can simplify the acoustic design of the handset.