[Noisebridge-discuss] Two-way optical interface?

[Ari Lacenski]

>
> Could any of y'all help me with that? I have no experience at all with
> photodiodes; I mainly do web apps, not hardware, so I don't even know
> where to begin there.
>

I'd be up for laying out my understanding of a plan of attack; other people
who've looked at analog components more recently could make better
recommendations on part numbers. Let me know off-thread if you want to
collaborate on that, I guess? It seems too much for one email.


> Also, any thought as to whether it'd be possible to run this through a
> single Thingy? I'm picturing eg little plastic stubs touching the
> actual board surface; those stubs underneath are connected to fiber
> optic, which has a splitter - one end taking input from individual
> LEDs, the other end going to a photodiode array for reading. But I
> have no idea if this kind of simultaneous i/o split is even possible.


The approach I have in mind involves multiplexing sensors that are all
being polled by a single Thingy. I've made 12 inputs at a time work. Not
19^2, but they needn't all be on the same processor.

<jazzhands>
There's another possibility in gaining an image of a whole board at a time,
using a wide-angle lens not a fiber-optic setup, and doing some edge
detection.. uh.. stuff . This is how the homebrewed multitouch table at
C-Base works, except it's using IR spectrum instead of visible. I know
basically nothing about camera image processing! Yay.
</jazzhands>

>> Or you could hack up a two-color monome with the buttons at crosses.

>That looks pretty neat equipment. Though expensive and sold out, and
not available in 19x19. ;-)

Sparkfun sells https://www.sparkfun.com/products/7835, which is readily
chainable, 97% as nice, and 100% as much fun to poke.

Ari

> Daylight might ruin this whole plan; choosing a photodiode responsive to
> > only a certain spectrum would help. I recommend NOT choosing the IR
> spectrum
> > to play with.
>
> *nod* Ideally I'd like it to work in a reasonably wide range of
> lighting conditions; the display part might need to be pretty bright
> to be seen though, especially when diffused as a halo.
>
> On the up side, if the stone is covering the Thingy/photodiode, then
> that blocks most outside light. Worst case scenario, it shouldn't be
> too hard to distinguish 'stone' from 'no stone'. I don't know if my
> "shoot light at it and read the reflectance through the same Thingy"
> idea is viable, but at least the presence of the stone creates a
> relatively controlled situation.
>
> Though hm, it wouldn't necessarily be *directly* covering the Thingy
> ('cause go stones have a relatively small contact surface, and move
> around a bit).
>
> This is also potentially a method for playing chess or the like, with
> the same board. You'd need to cover the pieces' bases with something,
> but then you get to control what that something is for optimal
> distinctiveness.
>
> > It might not work at all, but little cones of light spilling from the
> edges
> > of stones sounds like a lovely game.
>
> *nod* I imagine one could make all sorts of neat variants if this
> works. First getting it to work, though...
>
> > Or you could hack up a two-color monome with the buttons at crosses.
>
> That looks pretty neat equipment. Though expensive and sold out, and
> not available in 19x19. ;-)
>
> - Sai
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.noisebridge.net/pipermail/noisebridge-discuss/attachments/20120824/c97d8434/attachment-0003.html>