New on Anelok TV: capacitive sensing

Werner Almesberger werner at almesberger.net
Wed Apr 30 17:13:23 EDT 2014


Felix wrote:
> Seems more like a software
> filtering problem than a hardware one.

That's what we have field-upgradeable firmware for :-)

> Perhaps using the longest edged of
> anelok for the slider could help with this also :P

I'm thinking of giving it a length of about 45 mm. Depending on
how I hold it, I get different swipe lengths:

- thumb on the side: about 35 mm
  http://downloads.qi-hardware.com/people/werner/anelok/tmp/swipe-side.jpg

- thumb from above: about 30 mm
  http://downloads.qi-hardware.com/people/werner/anelok/tmp/swipe-top.jpg

- thumb flat, holding device diagonally: 40 mm and more, depending
  on how much I involve the wrist.
  http://downloads.qi-hardware.com/people/werner/anelok/tmp/swipe-diag.jpg

The piece of plastic I'm holding there is the battery cover of the
Ben, about 6 x 4 cm.

Hmm, playing with it, I just realized that the by far most natural
movement would be to move the thumb vertically, like this:

http://downloads.qi-hardware.com/people/werner/anelok/tmp/swipe-vertical.jpg

By the way, Joerg has suggested the following design for a proper
linear sensor:

http://wstaw.org/m/2014/04/30/plasma-desktopp31889.png

Line widths about 0.2 mm, so the whole thing would be about 5.4 mm,
which looks like a nice size.

> Is a probed design (see apple stuff)

Heh, that could of course mean that it's patented. Wouldn't patenting
the wheel be just like them ? :-)

But yes, a wheel would work, too. E.g., the wheel of my dead Samsung
X830 feels right and it is about the same size as the Anelok design
with wheel:

http://downloads.qi-hardware.com/people/werner/anelok/tmp/wheel-designs.jpg

One big benefit of a wheel pattern would be that I could conveniently
place the battery under it. Hmm ...

> Anyway, the rotary design from TI
> is think is better: http://imgur.com/yR3vvSJ

It certainly wins the style award :-) Generating that sort of pattern
looks tricky, though.

- Werner



More information about the discussion mailing list


interactive