There are several subtle differences in this description but it clearly represents some aspects of the original patents from AP.
What is clear is that it is an optimisation of the structures in AP's patents that describe the fundamental functionality of AP's original designs.
We know from earlier announcements that some of AP's early chips seemed to age and fail quickly.
"... Background technique A loudspeaker is a transducer device that converts electrical signals into acoustic signals. Loudspeakers are the basis for making audio equipment, acoustic active noise reduction equipment, etc. Therefore, the performance of loudspeakers has a key impact on the production of acoustic equipment. MEMS speakers (MicroElectroMechanicalSystem), that is, micro-electromechanical system speakers, have the advantages of good consistency, low power consumption, small size, and low price compared with traditional voice coil speakers.
At present, the commonly used MEMS loudspeakers realize analog sound by pushing the air through the movement of the diaphragm. Usually, the MEMS loudspeaker includes a vibrating plate, and the vibrating plate is provided with a plurality of dividing grooves arranged around it, and two adjacent dividing grooves overlap each other. A cantilever is formed, and the dividing groove divides the vibrating plate into successively connected diaphragm, cantilever and fixed part. With this structure, the diaphragm can perform reciprocating motion under the elastic action of the cantilever to simulate sounding. However, when the diaphragm moves up and down, it will drive The cantilever is bent and deformed relative to the fixed part. At this time, the connection position between the cantilever and the fixed part is in a stress concentration position, which is prone to breakage or plastic deformation, resulting in failure and affecting the service life of the speaker. ..."
These mods show solutions to problems that were only uncovered by building the original units and seeing their failure modes. It reflects some very smart re-engineering on 'somebody's' part, and most likely from someone who is very familiar with the physics of nanostructures as found in MEMS devices. That sounds more like an experienced MEMS fabricator than a customer. From AP's announcements this is likely to be the time that AP got involved with EM. I therefore imagine that this is how EM's input proved to be most valuable in that redesign process. This is likely the result. This also suggests that a clear collaboration has occurred.
I also note another feature. The patent describes the structure as 'vibrating' in its action. I am assuming that this refers to the ultrasonic base clock frequency of the natural oscillation of the structure. These vibrations are what are superimposed in DSR to create 'audible sound' waves of much lower frequency, as illustrated below:
![https://hotcopper.com.au/data/attachments/5413/5413930-ac6ba54bcfdcaec15295219b0f073273.jpg](https://hotcopper.com.au/data/attachments/5413/5413930-ac6ba54bcfdcaec15295219b0f073273.jpg)
The original designs had a top and bottom element that 'latched' the membrane at one or the other extreme of its travel range. It was serendipitously found with a partial structure that this may not have been not essential, and that good sound could still produced with only 'half' the structure. This led to the evolution of the current design.
But I imagine that this 'vibration' could also refer to analog vibrations at frequencies audible human hearing - i.e. 'sound'. This suggests other possible ways of deployment that could provide competitive products to the likes of USound etc, that are not based on DSR.
I am seeing some very interesting possibilities ahead for this partnership based on their patents!