Hacker News Reader

Comment by pg_bot | Hacker News Reader

Could you put a pressure sensor on the pad and only power on the coils directly underneath the object? I assume with software you could selectively turn on/off areas of the mat to manage the heat problems that come with multiple coils being on at once. I could see it being a problem if they were charging all at the same time but you should be able to fake it and call that good enough. All the devices have to charge, but they don't have to all charge at once. I think you could get tunnel vision and forget that.

Could someone who is better qualified tell me why that wouldn't work?

View on HN

There wouldn't be any need for pressure sensors; this already happens. The charged device adds some power to its own coil and the charging station detects that (it's unidirectional communication). For example my charging station lights up about a second after putting my phone down, and can show the battery status with the light. I would assume that when there's nothing on them the coils transmit in a low-power mode, with the protocol allowing the device to ask for more/less power as required (edit: Wikipedia confirms that the device can actually send some pretty precise instructions to the station, with voltage controllable in 50 mV increments [0]).

So logically it would seem the heat issue happens when there are three coils active at the same time transmitting at high power - the peak heat output is the issue.

[0]: https://en.wikipedia.org/wiki/Qi_(standard)

What I don't follow is how other multi-charging mats do exactly this. According to the article, by their estimation it was powerful harmonic EM interference that did AirPower in. If AirPower were only powering 3 coils at a time like other multi-charging mats why did it run into this problem when others had not

I guess it is because no other mat has overlapping coils. In the apple one they wanted you to not care where you put down your devices. Since they have so many coils, it then becomes likely that the 3 active ones overlap.

There has to be something else going on here. It strikes me as relatively simple to have the device talk to the mat and optimise on which coil was providing the most current draw.

I suspect that engineering does have a design that meets regulations but they couldn't get the cost down. I.e. the extra circuitry in both the met and possibly the devices to precisely sense which coil(S) are providing optimal charging is cost prohibitive.

Right. Or similarly, you could figure out which coil(s) were drawing the most power and perhaps power the others down?

I suspect they could do this but it turned out they couldn't get the cost of the design down.

So, they have a working product, but once they had a better idea of the numbers they killed it, hence the late stage cancelation.

Reply on news.ycombinator.com