- You are here:
- Home >
- Something for nothing (and the chips for free?)
Something for nothing (and the chips for free?)
– with apologies to Dire Straits
I’m a firm believer in the old maxim that “there’s no such thing as a free lunch”, especially when it comes to electronic systems (or government handouts) so it’s always good news when someone comes up with an idea where a single device can be used for two purposes.
We all know that silicon junctions are photosensitive. Back in the days of UV erasable chips, I once had a microcontroller that wouldn’t program. It turned out that sunlight steaming in through the window was turning on so many of the internal transistors that the programmer thought the chip was faulty as it was pulling so much supply current. Putting a sticker over the chip’s window fixed that one! Going even further back, older, wrinklier readers may remember scraping the paint from an OC71 to turn it into a phototransistor (or did my Grandad tell me about that one?)
The potential free lunch turns out that, not only can you can use LEDs as light emitters, you can also use them as photodiodes. The effect was publicised in the 1970’s by Forrest W. Mims, and resurrected by engineers at Mitsubishi in the USA in 2003. In their paper “Very Low-Cost Sensing and Communication Using Bidirectional LEDs”, Paul Dietz, William Yerazunis, and Darren Leigh present a method for multiplexing LEDs so that they can be used (almost) simultaneously as illuminators and sensors, and how to configure them as low-speed bi-directional data conduits.
Taking the concept forward further, Jeff Han, a Consultant at the Department of Computer Science, New York University, has developed an 8x8 LED panel used as a multi-touch pad type interface. There’s an excellent video demonstration of this in action.
The hardware is simple. In operation each LED is connected between two digital ports on a microcontroller (one output, one I/O) and forward biased to illuminate or reverse biased followed by a short “floating” period for sensing. The illumination level is determined by the amount of time the capacitance of the reversed biased diode (which is illumination dependent) takes to decay.
OK - so it’s not entirely free as you do need to use two ports per LED and there’s a slight software overhead, but it’s still a very interesting concept.
By Ray
