Apple cancels "AirPower" mat

[ebastler]

... since they couldn't quite get it to work:
https://techcrunch.com/2019/03/29/apple-cancels-airpower-product-citing-inability-to-meet-its-high-standards-for-hardware/

Apparently that "wireless power" thing is difficult -- even for inductive coupling over a short distance, when you only want to give the user a little freedom in positioning their gadgets. I am not holding my breath for the uBeam IPO...

[Raj]

  they had one job... Only one job... And they didn't finish it. All they had to do was to be of equal standard to the competition

[//Matt//]

Wow, they basically had to make a slim white version of the ten dollar Chinese wireless chargers. I don't see what could go wrong.

[GregDunn]

Heat, because too many coils too close together.  Most of the cheapo chargers only have a single coil.

I'm OK with Apple failing once in a while; it at least means they're trying to move stuff forward - unlike Microsoft, which is still struggling to produce a GUI which works better in 2019 than it did in 1985.

[//Matt//]

Heat, because too many coils too close together.  Most of the cheapo chargers only have a single coil.

Oh OK, that makes sense.

[EEVblog]

Anyone got links to the patents?
I only found this:
https://www.patentlyapple.com/patently-apple/2018/07/apple-reveals-more-technical-details-about-airpower-in-a-european-patent-filing.html
But they don't link or give a patent number

[EEVblog]

Heat, because too many coils too close together.  Most of the cheapo chargers only have a single coil.

Not quite sure I understand the thermal issue here. Surely only the nearest coil will be on, so what's the difference with a normal Qi charger coil?
Maybe some periodic sampling overhead, but not much?

[BrianHG]

ohhhhh, poor Apple...

[TopLoser]

Heat, because too many coils too close together.  Most of the cheapo chargers only have a single coil.

Not quite sure I understand the thermal issue here. Surely only the nearest coil will be on, so what's the difference with a normal Qi charger coil?
Maybe some periodic sampling overhead, but not much?

Multiple devices placed at random positions on the same mat was the goal.

[EEVblog]

Not quite sure I understand the thermal issue here. Surely only the nearest coil will be on, so what's the difference with a normal Qi charger coil?
Maybe some periodic sampling overhead, but not much?

Multiple devices placed at random positions on the same mat was the goal.

Yeah, but how does that create thermal issue?
It searched for devices and only turns on that coil fully when a device is over it. What's the difference between say two normal Qi coils side by side in the same case and this design?

[m98]

Probably one of those projects where you'd seriously underestimate the required effort to get it to work flawlessly. Surprisingly this one didn't work out even with the practically unlimited resources they have.
Maybe they couldn't make those layered coils work reliably with the Qi receivers.

[TheNewLab]

Anyone got links to the patents?
I only found this:

No product, no patent filed...? Chinese orginated from, never bother with filing patent?  Want to keep secret and never filed the patent? 
The last one would be dumb..I would guess anyone could reverse engineer the design..

Not quite sure I understand the thermal issue here. Surely only the nearest coil will be on, so what's the difference with a normal Qi charger coil?
Maybe some periodic sampling overhead, but not much?

Multiple devices placed at random positions on the same mat was the goal.

Yeah, but how does that create thermal issue?
It searched for devices and only turns on that coil fully when a device is over it. What's the difference between say two normal Qi coils side by side in the same case and this design?

Did I her 30 plus coils? powered from looked like a USB connector?? 5v and at best 2A? I wonder if it is not the coils themselves, rather the buck converters to power up those many coils?

Maybe Apple just did not get managing only the coils around the unit to power up only.  Even Apple has done some stuff that engineering just missed a solution right in front of their eyes...

[Raj]

Solution-Apple phone charging heated blankets 

on a serious note...they could have had a photodiode grid to detect the position and type of objects on it, and turn on a coil, that lines up the best...disconnect the rest so they don't cause cause any problem.

Butterfly keys is what they should have cancelled.

[Cyberdragon]

This is a company that failed at laptop thermal management in a released product, what do you expect?

[ebastler]

i think instead of searching in apple, search for samsung or others

https://patents.google.com/patent/US8629654B2/en
this one has what look like 28 coils wow
so apple wants to do 29 coils? haha 30? 31?

Well, in fairness: Having a patent on something does not mean that you got it to work. You can draw as many coils in your patent drawings as you like, whether or not you have a working prototype, and whether or not you ever intend to build such a version. I am sure Apple has patents on the AirPower charger as well.

[David Hess]

I saw what they based this on and the patent but did not keep the links.  The problem I saw was that the performance claims were based on scaling up the small signal performance of the prototype without considering efficiency which was poor because of added switching elements in series with the tank circuits.  Apple's demands for something thinner and lighter just made the situation worse.

I suspect the engineers at Unitrode who designed resonate switching power supplies could have done it or proved that Apple's requirements could not be met but I doubt Apple would have listened.

Could you separately drive high efficiency mutually coupled tank circuits?  It would be expensive and the resonate frequencies would not be identical so I am not sure how well it would work.  Resonate converters do something like this though.

[Zbig]

Not quite sure I understand the thermal issue here. Surely only the nearest coil will be on, so what's the difference with a normal Qi charger coil?
Maybe some periodic sampling overhead, but not much?

Multiple devices placed at random positions on the same mat was the goal.

Yeah, but how does that create thermal issue?
It searched for devices and only turns on that coil fully when a device is over it. What's the difference between say two normal Qi coils side by side in the same case and this design?

I think the idea was to have multiple coils overlapping, Olympic rings way. I imagine in this case you'd have a problem with coils shielding and/or partially shorting the active one even when they're themselves effectively disconnected?

[JohnG]

Another patent that may be relevant: https://patentimages.storage.googleapis.com/8f/fb/28/1a9fba81ff81bb/WO2018170282A1.pdf

John