Nanonote like clamshell based on AM335x processor

Werner Almesberger werner at almesberger.net
Sat Aug 4 14:49:02 EDT 2012


rakshat hooja wrote:
> From the little feedback I have received about the Nanonote, they really
> like the production quality of the product

Thanks for the flowers ! :-) I wasn't aboard Qi-Hw when they made
the Nanonote, but since I've heard much of the backstory, let me
try to fill in some details, as far as I think I know them:

There are three areas to consider: 1) the software, 2) the
electronics, and 3) the case.

"We", meaning the Qi-Hw community and closely cooperating projects
like Open Wrt, JLime, etc., can take full credit for the software,
i.e., the availability of a feature-rich Linux distribution that's
been adapted for the capabilities and needs of the Nanonote.

Regarding the electronics and the case, the Nanonote is derived
from an earlier product design, a Chinese dictionary, that Sharism
licensed. Some modifications had been made specifically for Qi-Hw,
such as the keyboard, and also some changes to the circuit, plus
other details like branded batteries.

We have Free schematics of the Nanonote but layout and case design
remain with the original company. The Nanonotes have also been
manufactured by the company that had produced the original
dictionaries.

I think this was a very smart approach, because it allowed Qi-Hw
to have a hardware platform to hack on in almost no time, allowing
the project and its community to learn and grow.

What didn't go so well is that we lacked (and still do) the
resources for a follow-on product that would overcome some of the
more bothersome shortcomings of the Ben.

> and are quite amazed how you got
> the cost as low as 99 usd and got such a good open product out for such a
> small run.

The low price is mainly due to hardware R&D having been written
off already. A product completely designed from scratch would
have to have a considerably higher price and/or larger volume
to recover the development and production setup cost.

> I did tell them that QI would most probably only focus on developing
> products whose production plan is fully sharable and hardware documentation
> is publicly availableI.

Yes, that would roughly be the common requirements. Part of
keeping things Free is also the use of Open Source tools, like
KiCad for EDA, and possible FreeCAD for mechanical designs.
Where we find such tools lacking, we extend or complement them.

I wouldn't be as harsh as Joachim in dismissing any piece of
hardware that has a bit of closed IP in it, but any such
subsystem would have to be considered as inexistent as far as
we're concerned.

I.e., the product would have to make sense even without using
that subsystem, be it 3D acceleration, wireless, or whatever.
Furthermore, that subsystem would have to be considered dead
weight, adding chip cost and possibly other issues at no benefit.

Opinions are divided where binary-only firmware that gets loaded
into some chip is concerned. As a pragmatist, I don't see them as
evil, as long as they can be freely redistributed, but I also
recognize their potential for causing trouble. More ideologically
inclined people would rather draw and quarter them, then burn the
parts at the stake, and hang the ashes :)

> My question to the list is how and if can we take this forward?

Well, the question is what TI are looking for. If they're
interested in developing a Nanonote-like device in cooperation
with the Qi-Hardware community, I could definitely see potential
there.

Within the Qi-Hw community, the abilities for making electronics
operating at moderare speeds and the system software that runs on
them should be more than adequate. We would need to find equipment
or help when it comes to debugging high-speed designs.

We lack experience in the area of mechanical engineering, though.
I.e., we can make some types of cases, e.g., the one for the
Milkymist One, but we currently couldn't pull a case as smooth as
the Nanonote's out of our hats. This could of course be an
opportunity for improving.

Since high-quality mechanical design is a fairly costly process
that requires specialized machinery, a reasonable goal could be
to make a fully open case design in prototype quality and have a
more polished but not easily reproducible variant made for actual
production.

If TI aren't looking for cooperation but just want to find a
low-cost clamshell case, perhaps Wolfgang could put them in touch
with the manufacturing company.

- Werner




More information about the discussion mailing list


interactive