Anelok: rfkill concept for the next board version

Werner Almesberger werner at almesberger.net
Sun May 11 12:56:31 EDT 2014


Ron K. Jeffries wrote:
> Your goal is not lowest price or smallest package.

Both things are actually pretty high on my list :-)

Lower production cost means that less upfront financing is needed
and also that there can be larger margins. This in turn leaves
more for further development, e.g., software improvements and
follow-on versions.

For example, it would be nice to develop a functionally equivalent
device with different chips, e.g., an STM32 MCU. If some day a
problem with the Kinetis surfaces, be it related to sourcing, some
technical weakness, or a successful security-related attack against
the hardware, one could quickly switch to that "plan B" design.

Larger margins also mean that there's a buffer in case something
turns out to be more expensive than planned or if price
expectations in the market change.


A small package means that the device will be easier to carry. I'd
like the size of the device to be determined mainly by the display.
The size of board 0 (with wheel) was largely defined by display and
wheel, though I had to add a few mm because I couldn't fit all the
circuit otherwise.

The next version will have some space-saving changes but there will
also be a few new components, mainly for more efficient or more
correct power handling. So I already know that space will be tight.
Any component I can kill off early is one problem less to solve :-)

And of course, fewer components mean less sourcing and manufacturing
risk, higher yield, probably more reliability in the field (something
that isn't there can't break), and so on.

- Werner



More information about the discussion mailing list


interactive