SparkFun Forums 

Where electronics enthusiasts find answers.

Have questions about a SparkFun product or board? This is the place to be.
By mutsop
#157534
Hi,

After a few small projects I'm trying to plan out a big one :)
And yes, it's another led table... It's been a dream though for a couple of years to have an interactive table (ever since the microsoft surface table was introduced).

So the idea is to create a tabletop of 12 x 18+ cathode RGB leds (216+). Each of these leds would be put in a halfenclosed box of 5x5cm

Now whats the cheapest solution to identify a touch? After some reading I came across IR proximity sensor, but it costs around 10€ a piece. (100+ is about 8€/piece), which seems a bit expensive If I need around 216 (1 for each box + some reserves).

Fyi I'll be using a semi diffused plexi plate to cover up the table.

So is there maybe another solution?
By Mee_n_Mac
#157546
A few other solutions come to mind. First be aware that LEDs can also be used as photo-diodes and so detect light impinging on them (w/some restrictions). So one way that's been done is to embed some IR LEDs as emitters and detected the reflected light of your hand via the other RGB LEDs.[ERROR - see posts below] This implies the RGB LEDs are off (unlit) during the detection process. I suppose you could also very briefly (so briefly it's not visible to the human eye) turn on some of the LEDs in your table and detect the presence of a hand via other LEDs (no need for the IR LEDs this way).

Then again if the table is in the daylight you might detect the shadow of a hand, blocking the light from overhead.

And yet another way, suggested here, was to use the capacitance of the LEDs as a detection method. LEDs have an inherent capacitance and it varies when you touch them, just like the metal plate capacitance detectors/switches. You need to find a way to measure the change in capacitance.

No doubt there are other ways. Search this forum and Google about a bit and you'll see some applications of the above ideas.
Last edited by Mee_n_Mac on Thu Mar 28, 2013 10:49 am, edited 1 time in total.
By jremington
#157560
LEDs used as photodiodes are wavelength sensitive -- similar to a resonant circuit. They will detect only light that is fairly close in wavelength (color) to the light they normally emit, so the idea of using an IR LED as an emitter and an RGB LED as a detector won't work.

The table idea is fun. You could also imagine using IR proximity sensors to get a distance reading as your hand approaches an area and change the color or the intensity of the light emission. See https://www.sparkfun.com/products/242 However, some covering materials, like diffused plexiglass, would probably cause the IR sensor to malfunction.
By Mee_n_Mac
#157571
jremington wrote:LEDs used as photodiodes are wavelength sensitive -- similar to a resonant circuit. They will detect only light that is fairly close in wavelength (color) to the light they normally emit, so the idea of using an IR LED as an emitter and an RGB LED as a detector won't work.
Yes and no. I've got it backwards in that an IR LED will detect photons of a higher energy/shorter wavelength. They work best at the same wavelength as they transmit but they will work in most cases. Red LEDs will detect green and blue, and green detect blue but RGB will not detect IR.

http://www.sensorsynergy.com/helpfulhints.htm
By jremington
#157581
I've measured the response curves of several different colored LEDs in photodiode mode and the curves do closely resemble that of a resonant LC circuit, except that the response peak is shifted somewhat in favor of shorter wavelengths from the emission maximum. Red LEDs would be extremely inefficient at detecting blue or green light, if at all. The "helpfulhints" page is misleading.

To illustrate this, below is an example of typical response curves for IR LEDs, taken from Forrest Mims' famous publication in Geophysical Research Letters V. 29, issue 13 (2002), in which he proposed to use LEDs to monitor various aspects of solar radiation spectra and atmospheric absorption:
LED_response.png
An interesting corollary is that if you wish to make an IR transceiver using IR LEDs as both emitters and detectors, it is important that the response curve and the emission curve match well. That is not the case for two identical LEDs.
You do not have the required permissions to view the files attached to this post.