Re: One for the Apple-ites!

[Ian Oliver <lists@xxxxxxxxxxxxx>]

In article
<f55662fc0911020103n77610019l5efc99ef10cec3ce@xxxxxxx>, Mark
McCall wrote:
> One of the problems with hubs and cheapo usb adapters is that they
> dont have the logic or the means to deliver 500mw. They're just dumb
> 100mw devices.

A few corrections.

USB initially provides a guaranteed 100mA and devices can then
negotiate additional helpings of 100mA up to a maximum of 500mA.

"Dumb" chargers do not do enumeration and negotiation, so the
base case
is that a device plugged into one of these will only take 100mA. Apple
devices, in their wisdom, often refuse to charge at all.

Obviously this is a serious limitation, and "something had to be
done".
What everyone has done is to define a new standard, and this specifies
that if a device detects that the USB D+ and D- lines are shorted
together and floating, they can take up to 1.3A from VBUS.

Of course, Apple have gone their own way. They need 2.5V on D+ and 2.0V
on D- before they will charge. I'm not sure what current they take if
they detect this.

The easiest way to provide this is -
5v---100k---(D+)----50k(2x100k in parallel?)----(D-)---100k-----0V

Any resistors will do, and some people supply 2.5V to both D+ and D-,
so experiment. A standard USB lead can be hacked to add the resistors
to make it into an Apple charger lead.

Ian Oliver
Sunny Leeds, UK
Using Java on Tini for control via Dallas 1-wire




------------------------------------