Post anelok

Werner Almesberger werner at almesberger.net
Fri Sep 20 07:44:07 EDT 2013


Ron K. Jeffries wrote:
> The hardware designed for Werner's anelok password safe will be interesting
> for other use cases. I hope it's possible to bring out SPI (and/or I2C)

In the current schematics draft (*), I've earmarked the five
unused pins for the four navigation buttons of the larger wheel
and a possible analog light sensor input.

(*) Still before layout, so it's likely that I'll move some pins
    around. I also have to check that I'm not doing something
    stupid then it comes to wakeup.

The two SPI units are already taken, for display and RF, but
I2C happens to be available on those unused pins. So if they
stay unused and don't get moved elsewhere, you could solder some
thin wires to these pads.

Likewise, you could implement bit-banging SPI on those pins. It
may also be possible to reuse the SWD (serial wire debug) pins,
which will be routed to pads as well (*). So that's two more.

(*) They're needed for "factory" tasks like flashing and testing.
    The idea is to have large pads for this sort of connection
    and have a fixture with a "bed of nails" (spring-loaded
    contact pins). Here's a home-made example of such a critter:

    http://downloads.qi-hardware.com/people/werner/tmp/lt-pgm-2.jpg
    http://downloads.qi-hardware.com/people/werner/tmp/lt-pgm-ben.jpg

> ability to use larger batteries e.g. AAA or AA would be nice.

That may actually be possible with the current design if you
rearrange/"patch" a few traces, to move the MCU and RF to the
regulated 3V3 rail and feed the regulator directly from the
battery.

The DC-DC boost converter is designed for this kind of use. A
lower battery voltage means that the converter has to use more
current. For Vbat = 1 V and Iout = 200 mA, I get 1.07 A. The
inductor I plan to use is rated at 1.1 A. So it may just work :)

Of course, if it doesn't, you get something like this:
http://people.openmoko.org/werner/burntmoko-inside.jpg

(That's what was left of an overloaded inductor.)

> the hardware design and the user interface and memcard access software
> might be leveraged (repurposed) to build a number of different gadgets.

Another approach would be to simply make a design variant at the
schematics and/or layout level. Then you're completely free as
far as subsystems, pin assignment, form factor, etc., are
concerned and can change as much or as little as you wish.

That's the beauty of openness - you don't have to "hack" the
device, you can modify it at the source.

- Werner



More information about the discussion mailing list


interactive