The Light-Emitting Diode

Transcript

0:00Introduction

Hello everyone! Let me start with a small puzzle. Silicon is the most common semiconductor around; most chips and ordinary diodes are made of it. But look at the little red indicator light on a charger or a television. That light is an LED, and it is not made of silicon. So why can't we make an LED from silicon? You will answer that yourself in a few minutes. The LED does come up in JEE questions, so let's begin.

0:30How an LED makes light

So, what exactly is an LED? The letters stand for light-emitting diode. It is a p-n junction, heavily doped, and it is used in forward bias. Okay? Now, in forward bias, electrons and holes are both pushed towards the junction. So tell me, what happens when they meet there? Yes, exactly, they cross the junction and recombine. An electron falls into a hole, and in doing so it drops across the band gap. So it must give that energy out somewhere. Right? So how much energy does each recombination give out? Think about how far the electron fell. That is correct, it is about the band gap. In suitable materials, this energy comes out as a photon, a tiny packet of light. On screen, h is Planck's constant and nu, the Greek letter, is the frequency, so h times nu is the photon's energy. The wavy equals sign means approximately, and E g, the E with a small g below it, is the band gap. So the photon energy is equal to, or slightly less than, the band gap. Clear? Now the shortcut that turns every LED numerical into one line. Lambda, the Greek letter, is the wavelength in nanometres, the n m in brackets. It is about twelve hundred and forty, over E g in electron volts, the e V in brackets. That twelve hundred and forty is just Planck's constant times the speed of light, in these units. So a bigger band gap means a shorter wavelength. Got it?

2:10Which colour? Choose the band gap

Alright, which colour will an LED give? Our eyes see light from about four hundred to seven hundred nanometres, okay? So what band gap do we need for visible light? Use the twelve hundred and forty rule at both ends. Let's check. Twelve hundred and forty over seven hundred gives about one point eight, and twelve hundred and forty over four hundred gives about three point one. So the band gap must lie between about one point eight and three electron volts. And notice, the red end, at seven hundred nanometres, needs the smaller band gap. Students often flip this, so be careful, okay? So we use compound semiconductors. Gallium arsenide phosphide, the formula on screen, has a band gap of one point nine electron volts, and it gives red light. Right? Now, gallium arsenide, without the phosphorus, has a band gap of one point four electron volts. Will we see its light? No, we will not, because one point four is below one point eight. Its wavelength is longer than seven hundred nanometres, so it gives infrared, which our eyes cannot see. With me? And now, our puzzle from the start. Silicon and germanium, the S i and G e on screen, are unsuitable for LED lights. Their band gaps correspond to the infrared, and in them the energy of recombination is given out mainly as heat, not light. That is why your red indicator light is not made of silicon. Clear?

3:43Using an LED: current, knee voltage, resistor

Now, how do we use an LED in a circuit? First, the brightness. The light intensity increases with the forward current, so more current means more light. Okay? Next, the knee voltage, the forward voltage at which the current really starts to flow. Which LED needs a larger knee voltage, red or blue? Think about the band gap. That is correct if you said blue. Blue light has a shorter wavelength, so it needs a bigger band gap, and so a bigger forward voltage. So the knee voltage increases from red to blue, and it is higher than for an ordinary silicon diode. Right? Now a practical question. What if you connect an LED backwards, across a twelve volt battery? Is that safe? It is not safe at all, and this is a classic trap. In reverse bias, an LED gives no light, and its reverse breakdown voltage is low, only about five volts. So twelve volts in reverse is likely to destroy it. That is why an LED must never be reverse biased. Got it? And in forward bias, we always put a resistor in series, to set the current. The supply voltage is capital V, and V F, the V with a small capital F below it, is the forward voltage that the LED drops. The resistor takes what is left, V minus V F, so R equals V minus V F, all over the current, capital I, okay?

5:08Example 1: LED with band gap 1.9 eV, wavelength and colour

Now, Example one, an easy one. An LED is made of a material with a band gap of one point nine electron volts. Find the wavelength and the colour of its light. Pause the video and try it first. Done? Okay, let's check. We use our shortcut: lambda is about twelve hundred and forty, over one point nine, in nanometres. Okay? Twelve hundred and forty over one point nine gives about six hundred and fifty three nanometres. So which colour is that? Yes, it is red, because six hundred and fifty three nanometres is near the seven hundred end of the visible range. And one point nine electron volts is exactly the gallium arsenide phosphide we met earlier. Clear?

5:54Example 2: least LED series resistor on 6 V, power wasted

And now Example two, a little harder. An LED drops one point eight volts and is rated at thirty milliamperes maximum, the m A on screen. Find the least series resistance on a six volt supply, and the power then wasted in the resistor. Pause the video and try it first. Done? Okay, let's check. So R is just six volts over thirty milliamperes, right? That is the trap, so be careful, because it forgets the LED completely. The LED takes its own one point eight volts, so the resistor gets only what is left: six minus one point eight, which is four point two volts. With me? Now the current. We use the maximum rating, thirty milliamperes, which is zero point zero three ampere in S I units. Okay? So R min, the R with min written small below it, is four point two over zero point zero three, which gives one hundred and forty ohms; the capital omega is the ohm. Why the least? Because any smaller resistor would let more than thirty milliamperes flow, which is above its rating. Clear? Finally, the power wasted in the resistor. P is the voltage across the resistor times the current: four point two times zero point zero three, which gives zero point one two six watt, or one hundred and twenty six milliwatts. Use only the resistor's four point two volts here, not the full six. Got it?

7:21Quick revision: 1240/Eg rule, visible band gaps, resistor

So, quick revision. An LED is a forward-biased p-n junction, and lambda in nanometres is about twelve hundred and forty over E g in electron volts. Visible light needs a band gap of about one point eight to three electron volts, so not silicon or germanium. It glows brighter with more current, and it must never be reverse biased. And the series resistor is V minus V F, over the current. Now go and try the practice problems on this page. You will find them easy. All the best!