How do “Smart Port” or "Power IQ" USB chargers work? Anyone?

[salil]

I’ve noticed an increasing number of high current multi-port USB chargers (for charging consumer electronics like phones, tablets, etc) entering the market this year from numerous companies.

Some of them have “smart ports” on them (Anker calls it “Power IQ”), which means the charger can auto-detect if an Apple, Android, Samsung Tablet, Kindle, etc device is plugged into that port and change the voltage on the data lines accordingly to signal to the device that it can draw the maximum possible current.

Here’s one of these products from the company “Anker”, they call it "Power IQ"
http://www.amazon.com/Anker®-40W-Desktop-Charger-Black/dp/B00GTGETFG/ref=sr_1_5?s=wireless&ie=UTF8&qid=1405314327&sr=1-5&keywords=smart+usb+charger

Here’s one from "EZOPower", they call them “Super Charger ports” :
http://www.amazon.com/EZOPower-Charger-Adapter-Extension-Samsung/dp/B00KJ0QV9K/ref=sr_1_78?s=wireless&ie=UTF8&qid=1405188185&sr=1-78&keywords=smart+charger+5+port

Does anyone know how this is being implemented?  Has there been a teardown of one of these chargers?

At first I thought it might be some proprietary thing Anker developed, because they were the only ones I could find on the market with “Power IQ” (evidently just a trademark kind of like Apple’s “retina display") for a while, but this year there’s loads of other companies with products that perform this function as well.

[andtfoot]

Possibly something like one of these?
http://www.ti.com/product/tps2513a/description

[amyk]

One possible scheme is explained here (page 18):
http://www.ti.com/lit/ds/symlink/tps2543.pdf

[josem]

One possible scheme is explained here (page 18):
http://www.ti.com/lit/ds/symlink/tps2543.pdf

That looks really dodgy. They are trying to detect an iPhone by checking for known behaviour, but if Apple change the behaviour it won't work.

That "known behaviour" is actually documented in Apple's MFi program,  it's not something reverse engineered by TI. It was created before the USB Battery Charging spec (with Dedicated Charging Ports that you suggest) even existed and has remained the same since then.

I like these datasheets from Maxim and find them useful to understanding USB charging:

http://datasheets.maximintegrated.com/en/ds/MAX14578AE-MAX14578E.pdf (especially figure 3 in page 16)

http://datasheets.maximintegrated.com/en/ds/MAX14617.pdf

[ajb]

Maybe they have a micro with USB host that can detect the vendor ID of the attached device. It then does a bus reset and sets the D+/D- lines appropriately.

Some time ago I considered building a universal charger that would do just this.  The hard part is accumulating and maintaining the list of devices and charging requirements.  As a bonus, though, it would allow charging of things like PS3 controllers, which for some idiotic reason need to be enumerated by a host to start charging.

[josem]

I'll say it again. Stick to the USB standard. If the device doesn't support it, the device is crap and it isn't your problem. If Apple or Sony products don't work with D+/D- shorted it's because Apple and Sony suck.

The problem is you're not only ignoring Apple and Sony, you're ignoring all devices that support fast charging modes >1.5A.
This includes devices like the Samsung Galaxy S4, Note, Tab etc..

Taking the word from Anandtech which is usually very reliable:

"When used with the bundled charger, the Galaxy S 4 recharges much faster than HTC's One, despite using a larger battery. Obviously the Galaxy S 4 will charge with any USB charger, but the charge time will simply be longer. Samsung uses a voltage divider and signals the presence of their own charger by sending 1.2-1.3 V across the D+ / D- pins, this is similar to what Apple does with 2.0 or 2.8 V across the pins for various USB chargers they've shipped over the years.This signaling is essentially Samsung's proprietary tablet charging signaling which they've employed on the Galaxy Note 2 and now SGS4, in fact the two use the same exact charger, so it's worth tossing out your old ones and getting the appropriate one to take advantage of the faster charging."

http://www.anandtech.com/show/6914/samsung-galaxy-s-4-review/2

If you can afford to ignore all these devices that "suck", well good for you, I guess.

[josem]

So the Texas chip and all the other fancy chargers that try to detect Apple and Sony devices won't work with a Samsung device anyway. Are you going to audit every single popular device on the market, and what about when new ones appear? It's the road to madness. It might even harm your product, because who wants to a buy a charger where the box says "iPhone 4 compatible!" when the iPhone 5 is out now? It looks old.

1.5A is pretty fast charging for most devices, except some of the larger tablets. I prefer Qi charging anyway.

I wouldn't worry very much about Apple changing their charging specs, they've been using the same resistor network method since the iPod moved to USB charging, 10 years ago. The values are now well known even outside MFi.

But do agree that getting devices like Samsung S4 etc to consistently pull over more than 750mA is a challenge as they try to be smart. Maybe smart about poor USB cables, underpowered chargers or maybe smart to Samsung's bottom line, who knows. But definitely too "smart" to the point of being almost unpredictable.

So on the charger side you have to be intelligent as well. That's why I recommended Maxim devices, for example the MAX14667 which clearly states "supports USB Battery Charger (BC) revision 1.2 specification compliant devices, including all major brand USB devices from Samsung, Nokia, Blackberry, LG, Apple devices, and Samsung Galaxy devices in autodetection 1A or autodetection 2A modes."

http://www.maximintegrated.com/en/products/interface/universal-serial-bus/MAX14667.html

They've figured out all this stuff. I think TI hasn't.

As for Qi.. well that's an even larger can of worms :-)