25 June 2004
Out of print books
I think it's helpful to recommend good references, but it might be
Barring that, I suggest your readers check out www.abebooks.com
for a great selection of out of print books. But shop - prices vary widely.
Editor's Note: In fact, abebooks.com is listed as a resource
on The Citizen Scientist Resource
LEDs as light detectors
Thank you for writing. I am very glad to learn about your interest in using light-emitting diodes (LEDs) as spectrally selective detectors.
While a high school senior in 1962, I first got the idea that light sensors should be able to double as light detectors. So I connected an automobile ignition coil to a cadmium sulfide photoresistor, switched on the power, and observed bright flashes of green light emitted by the semiconductor. The green flashes were distinctively different from the yellow flashes of an electrical arc.
While studying government (my major) in college, I found that certain silicon photodiodes can emit near-infrared radiation that can be detected by similar photodiodes. I managed to send modulated tones between such photodiodes. In 1971 I demonstrated the ability of many LEDs to detect light while experimenting with an optical fiber communication system. By placing a single LED at each end of the fiber, it was possible to send signals both ways through the fiber with only a single , dual purpose semiconductor device at each end of the fiber.
Regarding interface circuitry, an LED can be used just like a photodiode. The circuit shown here is a very simple transimpedance amplifier than can be made using many different kinds of common operational amplifiers. The gain (amplification factor) of this circuit equals the resistance of the gain (or feedback) resistor. Thus, a resistance of 1 megohm (1,000,000 ohms) will provide a gain of 1,000,000.
This circuit is designed to detect a dc (continuous) signal. For rapidly varying signals, such as a light beam modulated by voice or digital signals, a capacitor should be inserted between the LED and the inverting input of the amplifier.
The circuit shown here is taken from "Sun and Sky Monitoring Station Workbook" (Radio Shack, 2003, page 15) a manual I wrote to accompany the Sun and Sky Monitoring Station that I recently developed for Radio Shack.
You asked about using LEDs that emit different colors. The short answer is yes, LEDs of different colors generally respond to similarly colored light. Thus, a green LED that emits at 525 nm or so will have a peak spectral response of about 505 nm. An exception to the rule are the blue LEDs made from gallium nitride (GaN) and related chemistries. Blue LEDs may have a very sharp detection response near 368 nm. Green GaN LEDs may have the same or similar detection response. While the main response is sharply tuned, there is a broad shoulder of reduced response extending into the green.
The spectral tuning available by using different kinds of LEDs forms the basis for the Radio Shack Sun and Sky Monitoring Station, which uses four different kinds of LEDs to detect wavelengths of about 505 nm, 625 nm, 816 nm and 930 nm. This very broad range of wavelengths from very simple, inexpensive LEDs permits this instrument to measure photosynthetic radiation (PAR) and to detect aerosols and the total column water vapor.
I have written extensively about this topic, and you might find some references by doing a search on www.google.com or a similar search engine using my name and "LED," "photodetector," and similar terms. For papers and books I have written that discuss this topic, visit www.forrestmims.org and click on the Publications button. Sheldon Greaves has written a review of the Sun and Sky Monitoring Station that appears in this issue of "The Citizen Scientist." This review touches on the topic of using LEDs as detectors. I will try to write much more about this in a future article.
Copyright 2004 by Society for Amateur Scientists