USB Power Madness

I've been a proponent of using USB as a standard low voltage supply bus for almost 10 years now. And it has really happened. Today it is hard to find a newly designed electronics gadget that would not have the ability to be charged from USB socket. But just because all USB sockets look the same, it does not mean they have the same capabilities. Most of you are probably aware some chargers are capable of "fast" charging and some will take 10 hours or more.

We are putting the final design touches to our new flagship consumer device and we've decided to add the fast charging feature to the two external USB ports we have. Initially I thought it required just boosting the power supply output to about 2.5 Amps per port. I was wrong.

Digging deeper we found there has to be a special (passive) signalling implemented, something called USB DCP (Dedicated Charging Port). And there are two conflicting DCP standards: Apple and the rest of the world (nor really surprising, isn't it?). Fortunately there are solutions to this problem: replace the passive DCP signalling with active. Meaning put a microcontroller that will sense what kind of device is connected and will emulate the proper behavior. TI has a chip for that.

In our case the algorithm will be even more complicated. DCP is not compatible with USB data. So if something is connected to a port, we will first sense if it is a device our software can recognize and communicate with. If so, the port will operate as a standard USB port. Otherwise we will switch the port to DCP mode and hand over the negotiation to the TI chip. It seems such action is exactly what a user expects. Designing for success means often designing the system to behave exactly how users expect it to.