This is from the Technology page on AP's website, the transcribed talks with Danny and Yuval on 6/6/22, and from the Jul 22 4C update.
Based on the info from AP's website I think that some part of the MEMS driving software would normally reside and run on a separate external controller chip that interfaces and controls multiple chips to serialise and synchronise them with each other. I have found no other information from any announcements about this external controller chip. I could be wrong, but I presume therefore that the digital audio signal is simply presented and fed onto an interface on the consumer product circuit board, from which the controller reads it and sends it to however many MEMS chips are present to covert it into coherent sound in real time.
This sounds pretty amazing stuff, especially from a manufacturers perspective!
From the AP site / Technology
"... Audio Pixels speakers also offer unprecedented flexibility. In sharp contrast to conventional speaker technologies where the types and number of drivers, as well as the enclosure and its electronics all must be customized and endlessly fine-tuned to a specific application; Audio Pixels offers a single chip design that is modular and perfectly predicable and linear. Given that the total number of “pixels” is the only factor determining quality, frequency and sound volume, manufacturers and device designers need only to determine the target acoustic performance and select the appropriate
number of speaker chips, which are serialized using a single controller.The homogeneous construction of the chips simplifies and reduces the cost of customer design, integration, assembly and product inventory. It also permits the Company to optimize its pixels, the array, the fabrication and test processes. ..."
- "... What you see here ..... Let’s start with the picture on the upper right corner. That is our desktop engineering development environment. What you see here is a board, in that board there is a single chip. What differentiates that from the demonstration systems is that the ASIC is intentionally mounted external to the chip so that we can do debugging in development. This can be achieved in a portable demo because this type of system needs to be connected to our network by using a variety of different software packages that can run off a PC. It is basically a processor that runs the chip, sends the signals to the ASIC and the ASIC to the MEMS and the MEMS produces the sound. So it is critical that we receive chips that are packaged with the ASIC integrated inside. ..."
[from 29/07/22 Update & 4C]
- Product and Demonstration Development
-During the period all aspects of our product development environments have been redesigned / upgraded to best conform with the technical requirements of the MEMS Gen-1 (simplified structure). The enhanced development environments have permitted a number of algorithmic improvements that improve sound quality and expand the bandwidth, now measured to produce the full audible spectrum down to 20Hz; performance that is normally restricted to speakers that are orders of magnitude larger. This pioneering achievement for the first time in the history of sound reproduction, delivers a small form factor speaker that is capable of reproducing hi-quality sound throughout the full audible spectrum (and beyond).
- Receipt of chips packaged with our integrated ASIC driver was a major milestone achievement. As has been reported the proprietary ASIC driver among the other things, converts the available device voltage (from 2.3V to 24V) into the higher voltages required to operate and control the MEMS transducer. This achievement for the first time enables the company to “detach” from large bench power supplies as well as the PC controlling the chip (and the substantial wiring of both), to be used in its natural environment, run by the system in which the speaker will reside, such as the soon to be released headphone demo.
- - Simultaneously various other technical aspects of the platform have been implemented and validated using the 80-chip demo board. While not fully populated yet (due to chip shortage) this demo board has proven all the fundamental theories involving the use of multiplications of chips (whereby each doubling of the number of chips provides an uncompromised addition of 6 decibels of sound pressure).
- - Both of the aforementioned demo systems are functioning and rapidly approaching release date, with the key tasks remaining (beyond the receipt of additional chips) involving the full debugging of the transported code (from development to native code), the completion and incorporation of our latest algorithms that are being specifically optimized for the Gen-I simplified structure, as well as the accurate measurement of the performance of the completed demo chips / systems.
(20min delay)