Power engineers have been doing this for decades (a century?); estimating the temperature of a transformer or motor from a simple ODE to see if the relays need to trip. The parameters are derived from the electrical parameters of the transformer.
I'm not arguing the merits of this patent (see my other comments). I'm attempting to provide some background on how this messed up system "functions".
That said, this process is not "estimating" anything. It's taking high resolution direct measurements[0] of driver temperature with awareness of the frequency and amplitude of the audio signal (also factoring in crossover frequency) and doing a bunch of stuff (this is well outside my area of expertise) to essentially overpower the individual driver(s) well past their intended electrical and physical specifications.
If I ever ended up on a jury for this (I wouldn't - I would get dismissed after one question in voir dire) even I would have a tough time associating (and recognizing as prior art) what you described with this process.
[0] - I've since been corrected on this - it is an estimation (of sorts).
The "bunch of stuff" is precisely the part where estimation comes in. There is no direct temperature sensor to measure from, the only thing available is the voltage levels on the speaker. As the patent says it's a 2 level estimate: estimate of impedance from voltages -> estimate of temperature from modelling the speaker's physical parametrs as a function of the driving signal and impedance estimate.
With patents it's not the general idea that's being patented. The patent itself cites many earlier patents which operate around the same general idea just with slightly different implementations on how the control process is fed and how it reacts.
It also cites another patent (https://patents.google.com/patent/US20120020488A1/en) which implements control of a speaker via combination of impedance measurements of the speaker while it's being driven + a baseline ("binding") measurement of impedance + ambient temperature that is taken at device power up - ie generating a per device calibration curve.
From quickly staring at these, the Apple patents threads the needle by:
* Implementing speaker control using the output of a temperature model (which is fed by the resistance measurement), instead of "directly" operating on speaker resistance.
* Not implementing per-speaker calibration.
(Note! There are other cited patents that I didn't look at, this wasn't meant to be exhaustive, just illustrative... and just to generally satisfy an itch)
Appreciate the clarification - like I said this is well outside my area of expertise and my eyes glazed over pretty quickly in reading but at this point we're borderline in semantics.
Exactly. In this case it's a "process". Generally speaking my take is there's almost never entirely new process (or design or anything else) in any "invention" - humans have been at this for hundreds of years and there's almost nothing that's entirely new.
I don't think I've ever seen a filing without tons of references.
I suppose the non-semantics point of everyone in this chain is: there is nothing about this patent from Apple that explains why using a dynamic temperature/power limit model for the speakers is uncommon in laptops. They are neither the first nor last to get a patent to do this, why then should it be the "real" reason most laptops don't?
See my other comments about patents in practice. They don't matter until you end up in front of a jury (or settle with a troll, which is a different case). A jury comprised of 12 random people off the street, specifically selected by counsel to be as ignorant as possible on the subject matter at hand (easier to influence).
My original point - this isn't enough of a distinguishing feature to drive sales and revenue for any manufacturer to risk that.
Ok, you sold 1% more laptops because you have good speakers. Then Apple sues you and potentially wins damages on a larger percentage of ALL of your sales of that "infringing" product.
Not worth it, not even close. Not even worth it to try to circumvent and still risk getting sued.
I doubt e.g. Dell, HP, Lenovo, and others have been shipping >100 million laptops per year longer than this patent has been around with most, if not all, without the feature primarily because Apple was one of many (and not the last) to file a patent on their implementation of a common feature in the sound industry and they are now afraid of and upreprade for legal battles related to making laptops. More likely few just care about crappy laptop speaker sound enough to write a driver modeling each device and speaker combination for a significantly wider variety of shipping hardware. But I suppose it's impossible to know without asking every manufacturer.
They don't care because it doesn't impact sales enough to care. If some R&D person in these orgs took interest in this it would get smacked down by PMs, execs, etc because it doesn't matter and they don't care. That's all well before it even gets to legal where this patent would be discovered. With these factors compounded it doesn't happen - and Apple has a tiny edge on what is probably their core market/demographic anyway.
Hmm I don't think we're saying the same thing at all. I'm not saying patents are part of the reason companies don't care, I'm saying they didn't care before the patent and there is no reason to suspect the patent is the real reason they don't care since. If any part of the reasoning for why not has to do with the patent being there then it's not what I'm saying, even if it would sound similar.
Patents do matter, you can get a patent knocked out at several phases in litigation prior to actually arriving at a trial. That's not even assuming the defendant goes and gets a stay to IPR the patent.
I agree, but in this case the only innovation I see is applying a known solution to protect large inductor coils (transformers, motors) to protect smaller, inductor coils (speaker drivers).
Apple can patent their actual model all they want (for example by adding the driven air as a cooling term), but the idea, generally, of building a model to estimate the temperature is prior art.
Speaker engineers have been doing this for decades too. Not necessarily by the specific method chosen by Apple (didn't read the patent), or with laptop speakers, but dynamic control of signals to achieve greater loudness within a specific power rating has been a thing for a long time AFAIK. Here's an example:
Aside from applying it to speakers, whats novel?