IR triggered LED matrix

Hi all!

I am trying to build two mirrored LED drawing boards for a school project. Basically the idea is that whatever you draw on one instantly appears on the other and vice versa.

My first idea was to use a pixart sensor out of a wii remote to track a pen with an IR LED in it and map the XY position to an LED matrix and just mirror that onto a second matrix. But seeing as I've only got a couple of weeks to build this, and I'm a noob when it comes to both electronics and programming, my lecturer advised me to come up with something a little simpler..

I'm thinking I could also use IR sensors for each individual LED and just build something like: when sensor A is triggered, LED A is lit in both matrices. But to do this I would need sensors with a very low range and narrow angle to avoid the entire board lighting up as soon as the pen approaches it. Also I'm not sure if it's even possible to use the row-column-scanning principal to read from a matrix of IR sensors. Not to mention whether I should use photodiodes or phototransistors or something else..

If anyone would be so kind to help me figure out the best way to approach this project, I'd be very much obliged. Thanks in advance!

Cheers, Dan

my lecturer advised me to come up with something a little simpler..

Good advice, I think you are still thinking too complex. Getting an LED matrix of reasonable resolution is more than a couple of weeks work let alone linking them with sensors.

How about a radius arm pantograph for the drawing input with motors on a matching pantograph for the mirrored output?

It is the June 83 article I wrote which describes the basics here:- http://www.doc.mmu.ac.uk/STAFF/A.Wiseman/Acorn/BodyBuild/BB83.html

Grumpy_Mike: How about a radius arm pantograph for the drawing input with motors on a matching pantograph for the mirrored output?

Wow, thanks! But putting aside the fact that that's very very cool and I would love to build something like it; I'm afraid that it's not ideal for this particular project..

What I'm building is a prototype for a drawing tablet integrated into a tabletop. So the mechanics involved in this solution would be a bit too obtrusive. Also both tablets need to accept input and output data simultaneously as they will be operated by different users. And given the limited size of the tablets, the drawings need to only stay visible for a set amount of time.

Of course the prototype can be smaller or lower resolution than the proposed product.. Say I would go for an 8x8 matrix; would that really take me several weeks to set up? Don't get me wrong, I'm not questioning your expertise! But perhaps you were talking about a higher resolution ;)

Say I would go for an 8x8 matrix; would that really take me several weeks to set up?

It depends on how you do it, with a ready built matrix and an matrix driver chip then no but if you want to make it out of individual LEDs and have built in sensors then yes, there is a lot of wiring in an 8 X 8 matrix.

This project is an 8x8 matrix with switches in each matrix position and it took me a couple of months:-
http://www.thebox.myzen.co.uk/Hardware/Econo_Monome.html

The matrix definetly would be the hardest part, and adding infrared would not help, however it is actually possible to use a regular led as a sensor you would switch between input and output so fast that you don't see them blink then you would have the stylus with a led as the point, and when the led matrix detects the point of light it turns high during its output stage the coordinate of the led would just be sent to the other board to be turned on Depending on how big and high res you want it I would suggest breaking the board into smaller pieces individually controlled, so that speed is not a problem

Grumpy_Mike: This project is an 8x8 matrix with switches in each matrix position and it took me a couple of months:-

Hmm, is it the programming or the process of wiring it up which took you the most time? If it's mainly the wiring I could try winner10920's suggestion, using 2 ready built matrices.

winner10920: it is actually possible to use a regular led as a sensor you would switch between input and output so fast that you don't see them blink then you would have the stylus with a led as the point

I would probably have to devise a casing that ensures the LEDs can't detect light from neighboring LEDs, as I imagine they wouldn't all switch between detecting and emitting light at exactly the same time. But that seems doable.. But it does sound like the programming required to make this work might be quite a challenge!

[edit] This seems like an even nicer solution! http://cs.nyu.edu/~jhan/ledtouch/index.html If I can get the LEDs to turn on when they detect less light I wouldn't even necessarily need a stylus :) [/edit]

is it the programming or the process of wiring it up which took you the most time?

Wiring and physically making, the programming was almost trivial but then I have been at this game over 40 years.

Well that link said:-

More info coming soon, I promise.

and I suspect it is an old link. I did not have too much success when I tried that technique.

Does it have to be LED? What about a graphics LCD shield, and two pots rather like an old etch a sketch.