I'm all for helping you, but the thing is, this isn't how
you approach engineering. You look at all the requirements,
and then decide how the existing design meets those requirements,
and which parts have to be changed. Or whether, in fact, the
Colibri is so far removed from what you want to do, to be a
poor starting point.
Your question implies a change to the power distribution system.
As if, certain rails are not available to the thing you are
trying to design. You need to examine all the subsystems,
their power requirements. Also, why the designer of the
existing design, chose to do his power conversion and
distribution in a particular way. You make a list of all
the power requirements for the design, and *then* make
the call as to where to put boost or buck converters etc.
Maybe if other parts of the design need +5V, you need a
You'll save yourself a lot of work, if you study the design
first, and save schematic edits, for when all of your planning
is finished. Making a change here, and a change there, may not
cut it. And would be the mark of an amateur.
The AAT1275 implies you want to run a USB port, off a battery.
It seems a rather radical shift, and suggests you're about
to pepper us with 20 more, context free, component change
Also, I expect you'll soon be asking "can anybody pick out
a 2.2uH inductor for me?". The answer to that is "No".
I hate picking inductors for stuff. Page 11 of the datasheet,
goes through some of the technical issues. For the more
complex boost or buck switchers, some designers prototype
the design first, and optimize it (perhaps changing inductors,
if they don't like how the prototype is behaving). In this
case, since the output of the AAT1275 is not powering a part
of the main board, you could use the design itself as your
prototyping platform, paying for your mistakes with a PCB respin.
Some component manufacturers, make available evaluation boards,
and maybe you could get an eval board with the AAT1275 on it.
The advantage there, is you'd get to see the inductor they
chose, and also have a platform for testing the circuit before
using it. (Loading the thing up to 500mA, and seeing whether
it regulates properly.) Of course, an eval board costs money.
What happens with some switching converter designs, is the
designer poorly estimates the power requirements of the
load, and the switching converter is optimized for one
particular load point. Suddenly, the designer discovers
what a poor choice the thing is, because it is so close
to overload, or not being able to drive the load. I can't
speak to the AAT1275, and how well it does its job, but
adding a component like that to a design, increase the