Thanks for this Triin. Great find! This seems to be right up to date. Publication date 14/8/19. Application 7/3/19.
There are a few things in here that explain a some of the things mentioned or talked around in earlier announcements.
Here is a quick gleaning of some of the more pertinent ones. The numbering is the page number and section in the patent document. I have not tried to reproduce the equations here in HC!
p18 [0118]According to certain embodiments of the present invention, and unexpectedly, the amplitude of the moving elements used in any of the systems shown and described herein, may be much less, e.g. an order of magnitude less, than that predicted by conventional acoustic formulae.
The patent shows experimental specs relating to reproducing voice range for telecoms using the current 1024 pixel array chip Theoretical SPL was expected to be 90dBSPL. The patent shows the surprising ability of a 1000 pixel array, with a stroke length of 150µ, to produce Sound Pressure Level of 89.6dBSPL at its lowest frequency of 300Hz at 1m. That is pretty good agreement. Further experimental investigation of this with different stroke lengths revealed that a reduction down to as little as 10µ gave as much SPL as was expected from 100µ. The lowest frequency is used as the key value because it will have the lowest SPL. Higher frequencies will be better. Read the patent for more detailed info.
The following section talks about overcoming the charge trapping that was at the heart of the problems experienced over the last few years (over-voltage requirements, charge trapping, etc) This was all discussed in previous announcements.
p20 [0120] When a moving element and an electrode, provided in accordance with certain embodiments of the present invention, are latched by applying voltage, termed herein "latching voltage", therebetween, the latching force may gradually decrease e.g. due to charge trapping effects. In order to maintain effective latching, the latching voltage is typically gradually increased over time so as to maintain a desired level of latching force. The electrode layer may be coated with a thin layer of dielectric (insulator), e.g Silicon Nitride or Sapphire. The negative pole of a power supply may be connected to the moving element and the positive pole to the electrode. Voltage is then applied to charge the two elements. Typically, when the moving element and the electrode are in close proximity, the strength of the electrostatic field between their surfaces is so strong it rips (ionizes) the molecules trapped in the small air gaps between those surfaces. The released electrons accelerate towards the positive electrode. They hit the dielectric material at velocities and energy levels sufficiently high such that some of them penetrate and sink into the dielectric material. Due to the non-conducting nature of the dielectric material, the penetrating electrons become trapped in the dielectric, masking some of the positive charge of the electrode with their own negative charge, thus reducing the attracting latching force between the electrode and the moving element. This process may be difficult to reverse as dielectrics, being insulators, do not allow electrons to travel freely through them. This problem may be avoided by reversing the polarity of the applied voltage such that the positive pole is connected to the moving element and the negative to the electrode. The accelerated electrons then impinge upon the moving element and not the dielectric and since the moving element is conductive (or coated with a conductor) the excess charge can be easily dissipated and absorbed by the power supply.
p20 [0122] It is appreciated that the dielectric layer may be applied to the surface of the moving element rather than to that of the electrode, in which case the positive pole of the voltage is connected to the electrode and the negative to the moving element.
p21 [0123] There are also techniques known in the art that can overcome charge trapping. One example is coating the dielectric layer with a thin conductive layer. This layer dissipates the charge applied by impacting electrons. Since this conductor is in contact with the moving element, it short cuts the excess electrons to the moving element.
The patent provides information that I suspect may explain the comments in the latest announcement. The following suggests to me that it is the reproducibility of the fabrication processes in which every pixel may differ very slightly from the designed in characteristics. If the pixels were all exactly the same at the most minute levels the output would be at its most predictable. The sum of slight deviations “smears“ the overall output very slightly, but the more speaker elements there are the less this matters.
p22 [0137] According to the sound reconstruction method described herein, the generated sound pressure is proportional to the number of operating speakers. Different frequencies are produced by varying the number of speaker pulses over time. Unlike analog speakers, individual micro-speakers typically operate in a non-linear region to maximize dynamic range while still being able to produce low frequency sounds. The net linearity of the array typically results from linearity of the acoustic wave equation and uniformity between individual speakers.The total number of non-linear components in the generated sound wave is typically inversely related to the number of micro-speakers in the device.
It sounds like this could be what is being referred to in the most recent comments by Fred Bart:
“… While the devices performed consistently including the playing of speech and music, the acoustic output was measured to be below our targeted specifications. The root cause for such deficiencies in performance has been traced to slight anomalies that exist between the perfect world of computer simulation, and the imperfect world of fabrication and material behavior, imperfections that were only identifiable once the fabrication process consistently met our electromechanical specifications.
Together with the vendor a plan has been formulated to not only revert the specification back to compliancy, but also diminish the devices dependencies on the vendor’s fabrications tolerances….”
All up I am reassured that they are clearly at the final stages of protecting the IP and in getting the manufacturing specifications tied down tight.
This is all good.
NB: note that all these patent docs are available to the public like us, so are also available to anyone in the industry!
(20min delay)