PDA

View Full Version : Demystifying LEDs.



xmastree
04-26-2006, 09:58 AM
LEDs, they're not as difficult to understand as some people make out.

The first thing to note is that there are two main parameters of an LED. The forward current and the forward voltage. This voltage seems to cause a lot of confusion. You apply a voltage, and you get current, right?
Well, that's right for a resistor, which is a linear device. In other words, current is proportional to voltage. The basic equation for calculating Voltage, Current and Resistance is V=IR, and we'll use that later.

However, unlike resistors, LEDs are not linear so they behave differently. Rather than applying a voltage and getting a current, you apply a current and get the voltage. It may sound the same, but it's not.

Ok, here's some maths. Not complicated, if you can add, subtract, multiply and divide, you can do this.

Consider a 150 ohm resistor with 3V across it. The current passing through it will me 20mA
V=IR therefore I=V/R, 3/150 = 0.02
That's simple. Now, reduce that to 1.5V and you'll get 10mA. Increase to 4.5V and you'll get 30mA

An LED on the other hand, rated 3V@20mA may well pass 20mA with 3V applied, but at 1.5 you'll get nothing and at 4.5V you'll probably burn it out.

So, let's try forcing a current through our resistor. Push 20mA through it, V=IR, .02x150, 3V
At 10mA the voltage is 1.5V and at 30mA it's 4.5V. That's linear.

The LED however, will give you 3V at 10mA, 20mA and 30mA (there will be a slight difference due to the internal ersistance, but it's negligible). So, if any current will give you 3V, what will 3V give you? Could be anything.

Why is this? Well,the semiconductor junction requires a certain voltage before it will conduct. This is called the breakover voltage. Once it starts conducting, it will try to maintain that voltage. Therefore, if you apply 4.5V to it, it will try to pass enough current to reduce that supply to 3V and probably burn out in the process.

So, to control the current through the LED you need a series resistor.

Using our 3V@20mA LED, to run it from a 12V supply you need to find the correct value for the series resistor. There are tools to calculate it, but it's really, really easy. We know the LED will have 3V across it, so the resistor will have the remaining 9V across it (3+9=12).
Using our old friend, V=IR we can make R = V/I, R = 9/0.02, R = 450 ohms. Since 450 isn't a preferred value, use the nearest which will be 470. Close enough.

Told you it was easy. 8)

It's not a good idea to put LEDs in parallel, since if there's a slight difference in the forward voltage the one with the lower voltage will take all the current. Why? Well, let's say that one gives 2.95V and the other is 3.05V. Since there are two in parallel, you'll want 40mA so that each gets its 20. So, once they're conducting, the voltage across them will be 2.95, since that's the lower one. That's not enough for the 3.05V one to start conducting, it needs 3.05V. In reality, the internal resistance will even things out a little, but it's not a good idea.

They can be used in series though. Two 3V@20mA LEDs in series will behave like one 6V LED. You can even use different ones providing they all require the same current. Just add the voltages together, subtract it from the supply and put that value in the equation for the resistor.

For example, two green LEDs, 2V@20mA, and one blue, 5V@20mA gives a total of 9V. Remaining voltage is 3V, so the resistor will be 150 ohms. The nice thing about this, especially for battery powered equipment, is that the total current drain for three is the same as for one. Simply put, the battery will last three times as long.

http://www.lotechdesigns.com/host/images/3156ledseries.png

crenn
04-26-2006, 11:29 PM
Hmmm.... very nice piece of writing. But you've forgotten that batteries and some PSUs aren't perfect and has a resistance as well. And your fact about the battery lasting longer isn't 100%, but yes, it will last longer. Anyway, I'll keep my mouth shut, good information though.

xmastree
04-27-2006, 01:35 AM
Hmmm.... very nice piece of writing. But you've forgotten that batteries and some PSUs aren't perfect and has a resistance as well.Yeah, but I don't wish to confuse things. This is aimed at computer modders, who will mostly be using the 12V supply. Whilst it has resistance, it's negligible so I didn't mention it.

XcOM
04-27-2006, 10:25 AM
my pro mod on the front of my machine is cool, i took the front off, removed the LED (Green) and replaced it with a blue one, being pro, and couldn't b arsed to medd about, i just soldered it to the old wire for the power LED< therefor burning it out (Was an LED i had lying about, couldn't let it go to waste)

So, new LED, this time i got a 3V DC LED with resister included for various voltages, works a treat now.

xmastree
04-27-2006, 10:40 AM
That's strange. Usually the power LED is just a regular LED and the resistor is on the motherboard. I've done exactly the same and it worked just fine. :?

crenn
04-27-2006, 11:18 PM
He may have hooked up the LED in reverse bias... although I don't think it would have enough power to blow an LED.....

But also LEDs have different tollerances of voltages for currents. There may have been the difference which caused it to burnout. And there is a resistor (as far as I know) on the motherboard for the front panel LEDs.

DaveW
04-28-2006, 08:57 AM
You havn't mentioned that LED's are one way! Of course, if you didn't know that then you shouldn't be mucking around, but everyone has to start somewhere.

-Dave

xmastree
04-28-2006, 09:44 AM
True, and true. Except some special ones which are basically two LEDs connected in parallel, one each way. Usually a red and a green. Current one way and it's red, the other way and it's green. AC, and they both light, giving yellow.

Anyway, since we're onthe topic, how to identify which way to connect them.

An LED, like any other diode, has two connections. They are the anode and the cathode. The current flows in the anode and out the cathode. So, the anode must be connected to the positive and the cathode to the negative. If they're in series, they should be anode to cathode, just like batteries in series go positive to negative.

The anode is usually the longer wire, and the cathode is usually marked with a flat edge on the package. The wire lenght isn't reliable if the LED has been used before, as the leads might have been cut. The flat on the package is more reliable, but not 100%. I have seen them with a flat to mark the anode, but not lately so maybe those manufacturers got what they deserve for flying in the face of convention for no good reason.

The method I always use is to look inside the LED at the internal construction. There are two things in there. One larger cup, containing the chip itself, and a smaller post which has a tiny wire connecting to the chip.

The CUP is the CATHODE.

In this photo:
http://www.lotechdesigns.com/host/images/5268m4.jpg
The cathode is the lower one, the anode post isn't so clear as it didn't catch the light, but you can still see it.

Here's a diagram instead:
http://i50.photobucket.com/albums/f341/chrispollard/LEDpolarity.png

Hmm, think I'll try some extreme closeup photography... watch this space.

xmastree
04-28-2006, 10:04 AM
http://www.lotechdesigns.com/host/thumbs/5447blueled.jpg (http://www.lotechdesigns.com/host/images/5447blueled.jpg)

The cathode's at the top in that photo. And, for those interested, here (http://www.lotechdesigns.com/host/images/427663RD0009.jpg) is how I took it.

Here (http://en.wikipedia.org/wiki/Image:Ledmrp.jpg) is a better one, but it's not mine.

jdbnsn
04-28-2006, 12:58 PM
I think this was a great write up, I learned alot about these things and you saved me a lot of time googling info. Thanks!

Guttenaffe
04-28-2006, 07:10 PM
jdbnsn: careful googling about LEDs over 1/2 the stuff I've seen on LEDs is flat out wrong. This article is 100% correct :) I have a few posts with this same info in it on this board.

xmastree:You stole my story :P I been writing a larger and longer article on LEDs on and off for a bit, though I will still finish it I got some cool animations and what not I made for it. I've been too busy with school and work to finish it. :/

I'd like to add the following:
LED's arenít the same as normal diodes as their reverse breakdown voltage is WAY lower (which can cause problems though we electronic engineers normally just put a Si, or Ge if need be, diode in series to prevent damage to a LED.)

xmastree
04-28-2006, 09:02 PM
jdbnsn: careful googling about LEDs over 1/2 the stuff I've seen on LEDs is flat out wrong. This article is 100% correct :) 8)


xmastree:You stole my story :P I been writing a larger and longer article on LEDs on and off for a bit, though I will still finish itGreat. More information is always good, so long as it's not conflicting. Since you agree 100% with what i wrote, I guess that won't be a problem. I'm interested to see your animations though...


I'd like to add the following:
LED's aren’t the same as normal diodes as their reverse breakdown voltage is WAY lower (which can cause problems though we electronic engineers normally just put a Si, or Ge if need be, diode in series to prevent damage to a LED.)True, they aren't as tough as rectifier diodes, but they're not meant to be either. As for putting a diode in series, isn't it better to put one in inverse parallel? The resistor will still control the current and you won't get more than .6V across the LED in reverse.

xmastree
04-29-2006, 06:20 AM
There's a good write-up on wikipedia, but it's a bit too technical for most users, as it goes on about which chemicals produce which light etc. I concentrated on their use, since that's what concerns us modders more than the fact that the newest method used to produce white light LEDs is based on homoepitaxially grown zinc selenide (ZnSe) on a ZnSe substrate which simultaneously emits blue light from its active region and yellow light from the substrate.

Erm, yeah... right... :?

The best bit of info there, is this line:
"Because the voltage versus current characteristics of an LED are much like any diode, they can be destroyed by connecting them to a voltage source higher than their turn-on voltage."

crenn
05-01-2006, 07:39 PM
Hmmm... you should also mention there are some very high powered LEDs and also ones which can take upto 6V due to the fact they have a small IC in them and 3 coloured LED chips.

Guttenaffe
05-07-2006, 08:05 PM
crenn, there are LED "bulbs" that require 3A to light and 5A for max brightness. There are all kinds of LEDs out there. This was just an overview of "general" leds.

crenn
05-07-2006, 08:54 PM
True, very true. There are a lot of different types of LEDs, I'm not denying it, but most of the time, the brightness is dependant on the current going through it.

xmastree
05-07-2006, 09:14 PM
Yep.

This wasn't meant to be an exhaustive article covering every available option. It was mainly meant to cover 'standard' types which most of us will use for our mods. Typically, running 10-20mA LEDs from a 12V supply.
Covering the more specialised types would probably make it more confusing so I didn't go there.

I mainly wanted to help people understand how an LED needs a certain current rather than a voltage, and how to achieve that. Once you understand that concept (which I admit, can be difficult to grasp) the rest is easy.