Light emitting diode




Diodes, like all semiconductor devices, are governed by the principles described in quantum physics. One of these principles is the emission of specific-frequency radiant energy whenever electrons fall from a higher energy level to a lower energy level. This is the same principle at work in a neon lamp, the characteristic pink-orange glow of ionized neon due to the specific energy transitions of its electrons in the midst of an electric current. The unique color of a neon lamp’s glow is due to the fact that its neon gas inside the tube, and not due to the particular amount of current through the tube or voltage between the two electrodes. Neon gas glows pinkish-orange over a wide range of ionizing voltages and currents. Each chemical element has its own “sinature” emission of radiant energy when its electrons “jump” between different, quantized energy levels. Hydrogen gas, for example, glows red when ionized; mercury vapor glows blue. This is what makes spectrographic identification of elements possible.
Electrons flowing through a PN junction experience similar transitions in energy level, and emit radiant energy as they do so. The frequency of this radiant energy is determined by the crystal structure of the semiconductor material, and the elements comprising it. Some semiconductor junctions, composed of special chemical combinations, emit radiant energy within the spectrum of visible light as the electrons change energy levels. Simply put, these junctions glow when forward biased. A diode intentionally designed to glow like a lamp is called a light-emitting diode, or LED.
Forward biased silicon diodes give off heat as electron and holes from the N-type and P-type regions, respectively, recombine at the junction. In a forward biased LED, the recombination of electrons and holes in the active region in Figure (c) yields photons. This process is known as electroluminescence. To give off photons, the potential barrier through which the electrons fall must be higher than for a silicon diode. The forward diode drop can range to a few volts for some color LEDs.
Diodes made from a combination of the elements gallium, arsenic, and phosphorus (called gallium-arsenide-phosphide) glow bright red, and are some of the most common LEDs manufactured. By altering the chemical constituency of the PN junction, different colors may be obtained. Some of the currently available colors other than red are green, blue, and infra-red (invisible light at a frequency lower than red). Other colors may be obtained by combining two or more primary-color (red, green, and blue) LEDs together in the same package, sharing the same optical lens. For instance, a yellow LED may be made by merging a red LED with a green LED.
The schematic symbol for an LED is a regular diode shape inside of a circle, with two small arrows pointing away (indicating emitted light), shown in Figure

LED, Light Emitting Diode: (a) schematic symbol. (b) Flat side and short lead of device correspond to cathode. (c) Cross section of Led die.
This notation of having two small arrows pointing away from the device is common to the schematic symbols of all light-emitting semiconductor devices. Conversely, if a device is light-activated (meaning that incoming light stimulates it), then the symbol will have two small arrows pointing toward it. LEDs can sense light. They generate a small voltage when exposed to light, much like a solar cell on a small scale. This property can be gainfully applied in a variety of light-sensing circuits.
Because LEDs are made of different chemical substances than silicon diodes, their forward voltage drops will be different. Typically, LEDs have much larger forward voltage drops than rectifying diodes, anywhere from about 1.6 volts to over 3 volts, depending on the color. Typical operating current for a standard-sized LED is around 20 mA. When operating an LED from a DC voltage source greater than the LED’s forward voltage, a series-connected “dropping” resistor must be included to prevent full source voltage from damaging the LED. Consider the example circuit in Figure using a 6 V source.

Setting LED current at 20 ma. (a) for a 6 V source, (b) for a 24 V source.
With the LED dropping 1.6 volts, there will be 4.4 volts dropped across the resistor. Sizing the resistor for an LED current of 20 mA is as simple as taking its voltage drop (4.4 volts) and dividing by circuit current (20 mA), in accordance with Ohm’s Law (R=E/I). This gives us a figure of 220 Ω. Calculating power dissipation for this resistor, we take its voltage drop and multiplyby its current (P=IE), and end up with 88 mW, well within the rating of a 1/8 watt resistor. Higher battery voltages will require larger-value dropping resistors, and possibly higher-power rating resistors as well. Consider the example in Figure (b) for a supply voltage of 24 volts:
Here, the dropping resistor must be increased to a size of 1.12 kΩ to drop 22.4 volts at 20 mA so that the LED still receives only 1.6 volts. This also makes for a higher resistor power dissipation: 448 mW, nearly one-half a watt of power! Obviously, a resistor rated for 1/8 watt power dissipation or even 1/4 watt dissipation will overheat if used here.
Dropping resistor values need not be precise for LED circuits. Suppose we were to use a 1 kΩ resistor instead of a 1.12 kΩ resistor in the circuit shown above. The result would be a slightly greater circuit current and LED voltage drop, resulting in a brighter light from the LED and slightly reduced service life. A dropping resistor with too much resistance (say, 1.5 kΩ instead of 1.12 kΩ) will result in less circuit current, less LED voltage, and a dimmer light. LEDs are quite tolerant of variation in applied power, so you need not strive for perfection in sizing the dropping resistor.
Multiple LEDs are sometimes required, say in lighting. If LEDs are operated in parallel, each must have its own current limiting resistor as in Figure (a) to ensure currents dividing more equally.

However, it is more efficient to operate LEDs in series (Figure (b)) with a single dropping resistor. As the number of series LEDs increases the series resistor value must decrease to maintain current, to a point. The number of LEDs in series (Vf) cannot exceed the capability of the power supply.

Multiple series strings may be employed as in Figure > (c).
In spite of equalizing the currents in multiple LEDs, the brightness of the devices may not match due to variations in the individual parts. Parts can be selected for brightness matching for critical applications.

Multiple LEDs: (a) In parallel, (b) in series, (c) series-parallel
Also because of their unique chemical makeup, LEDs have much, much lower peak-inverse voltage (PIV) ratings than ordinary rectifying diodes. A typical LED might only be rated at 5 volts in reverse-bias mode. Therefore, when using alternating current to power an LED, connect a protective rectifying diode anti-parallel with the LED to prevent reverse breakdown every other half-cycle as in Figure (a).

Safely driving an LED with AC: (a) from 24 VAC, (b) from 240 VAC.
If the LED is driven from a 240 VAC source, the Figure (a) voltage source is increased from 24 VAC to 240 VAC, the resistor from 1.12 kΩ to 12 kΩ. The power dissipated in the 12 kΩ resistor is an unattractive 4.8 watts.

  P = VI = (240 V)(20 mA) = 4.8 watt

A potential solution is to replace the 12 kΩ resistor with a non-dissipative 12 kΩ capacitive reactance. This would be Figure (b) with the resistor shorted. That circuit at (b), missing the resistor, was published in an electrical engineering journal. This author constructed the circuit. It worked the first time it was powered “on,” but not thereafter upon “power on”. Each time it was powered “on,” it got dimmer until it failed completely. Why? If “power on” occurs near a zero crossing of the AC sinewave, the circuit works. However, if powered “on” at, say, the peak of the sinewave, the voltage rises abruptly from zero to the peak. Since the current through the capacitor is i = C(dv/dt), the current spikes to a very large value exceeding the “surge current” rating of the LED, destroying it.
The solution is to design a capacitor for the continuous current of the LD, and a series resistor to limit current during “power on” to the surge current rating of the LED. Often the surge current rating of an LED is ten times higher than the continuous current rating. (Though, this is not true of high current illumination grade LED’s.) We calculate a capacitor to supply 20 mA continuous current, then select a resistor having resistance of 1/10 th the capacitive reactance.
  I = 20 mA
  Xc = (240 V) / (20 mA) = 12 kΩ
  Xc = 1/2πfc
  C = 1/2πXc = 1/2π60(12 kΩ = 0.22 µF
  R = (0.10)Xc= (0.10)(12kΩ) = 1.2 kΩ
  P = I2R = (20 mA)2(1.2 kΩ) = 0.48 watt
The resistor limits the LED current to 200 mA during the “power on” surge. Thereafter it passes 20 mA as limited by the capacitor. The 1.2 kresistor dissipates 0.48 watts compared with 4.8 watts for the 12 kΩ resistor circuit.

What component values would be required to operate the circuit on 120 VAC? One solution is to use the 240 VAC circuit on 120 VAC with no change in component values, halving the LED continuous current to 10 mA. If operation at 20 mA is required, double the capacitor value and halve the resistor value.
The anti-parallel diodes in Figure can be replaced with an anti-parallel LED. The resulting pair of anti-parallel LED’s illuminate on alternating half-cycles of the AC sinewave. This configuration draws 20 ma, splitting it equally between the LED’s on alternating AC half cycles. Each LED only receives 10 mA due to this sharing. The same is true of the LED anti-parallel combination with a rectifier. The LED only receives 10 ma. If 20 mA was required for the LED(s), The capacitor value in µF could be doubled and the resistor halved.
The forward voltage drop of LED’s is inversely proportional to the wavelength (λ). As wavelength decreases going from infrared to visible colors to ultraviolet, Vf increases. While this trend is most obvious in the various devices from a single manufacturer, The voltage range for a particular color LED from various manufacturers varies. This range of voltages is shown in Table .
Optical and electrical properties of LED’s

LED λ nm (= 10 -9m) Vf(from) Vf (to)
infrared 940 1.2 1.7
red 660 1.5 2.4
orange 602-620 2.1 2.2
yellow, green 560-595 1.7 2.8
white, blue, violet 3 4
ultraviolet 370 4.2 4.8

As lamps, LEDs are superior to incandescent bulbs in many ways. First and foremost is efficiency: LEDs output far more light power per watt of electrical input than an incandescent lamp. This is a significant advantage if the circuit in question is battery-powered, efficiency translating to longer battery life. Second is the fact that LEDs are far more reliable, having a much greater service life than incandescent lamps. This is because LEDs are “cold” devices: they operate at much cooler temperatures than an incandescent lamp with a white-hot metal filament, susceptible to breakage from mechanical and thermal shock. Third is the high speed at which LEDs may be turned on and off. This advantage is also due to the “cold” operation of LEDs: they don’t have to overcome thermal inertia in transitioning from off to on or vice versa. For this reason, LEDs are used to transmit digital (on/off) information as pulses of light, conducted in empty space or through fiber-optic cable, at very high raes of speed (millions of pulses per second).
LEDs excel in monochromatic lighting applications like traffic signals and automotive tail lights. Incandescents are abysmal in this application since they require filtering, decreasing efficiency. LEDs do not require filtering.
One major disadvantage of using LEDs as sources of illumination is their monochromatic (single-color) emission. No one wants to read a book under the light of a red, green, or blue LED. However, if used in combination, LED colors may be mixed for a more broad-spectrum glow. A new broad spectrum light source is the white LED. While small white panel indicators have been available for many years, illumination grade devices are still in development.
Efficiency of lighting

Lamp type Efficiency lumen/watt Life hrs notes
White LED 35 100,000 costly
White LED, future 100 100,000 R&D target
Incandescent 12 1000 inexpensive
Halogen 15-17 2000 high quality light
Compact fluorescent 50-100 10,000 cost effective
Sodium vapor, lp 70-200 20,000 outdoor
Mercury vapor 13-48 18,000 outdoor

A white LED is a blue LED exciting a phosphor which emits yellow light. The blue plus yellow approximates white light. The nature of the phosphor determines the characteristics of the light. A red phosphor may be added to improve the quality of the yellow plus blue mixture at the expense of efficiency. Table compares white illumination LEDs to expected future devices and other conventional lamps. Efficiency is measured in lumens of light output per watt of input power. If the 50 lumens/watt device can be improved to 100 lumens/watt, white LEDs will be comparable to compact fluorescent lamps in efficiency.