blue LEDs and resistors

[Servanya]

Howdy all.
I am having issues getting blue LED's to work on my computer. I tried replacing my power on light with a blue one, but it didn't work. I found a few sited using google that indicated blue and white LEDs require more voltage (about +4v), and my LED power supply was only giving about 2v. So, after poking aroung my motherboard, I found +5v coming out of my USB extender port (the place where you plug in the extra 2 USB ports that come with motherboard). So, I took an analog audio cable from my cdrom and plugged it in - I plugged the blue LED into the other end and it worked! Woohoo!

So - I was reading about resistors. What if I want multiple blue lights? Can I take the power from my power supply (+12v) and use a resistor? The table on the back of the LED package was confusing Is there any way to make my hdd light have a blue light? I would have to plug it into the +2v connector from the mobo so it knows when to blink - but there isn't enough power to light the darn LED if I do that!

For the resistors, on the back of the package (the LED package), it doesn't say what wattage to use, it just says 680ohm for 12v DC. I think I am missing something here. Any clarification for my cloudy mind would be greatly appreciated.

 

[phalanx]

hmm, that's a bit of pickle you got there
What I would do is use the +2V to drive an NPN transistor which in turn drives your LED.
That way you can use the +12V to power the LED, while using the +2V to drive it.
If you're not into electronics you probably don't have a clue how to use a transistor.
So here's a small schematic that will -hopefully- clarify it
I'll leave the implementation of this thing in your case to you

 

[Servanya]

Wonderful! I knew there had to be a way. I was thinking something similar - and the transistor is all I need. Very cool. Thanks for the schematic too.
A recent developement: I was just talking to a co-worker who did the same thing with the blue light- but he just plugged them in and it worked fine. Some people have all the luck.
Anyhoo--muchas gracias!

 

[StandardCell]

<< So - I was reading about resistors. What if I want multiple blue lights? Can I take the power from my power supply (+12v) and use a resistor?
>>



That's exactly how you'd do it. Mind you, if it's just power on, you have a different setup than for signals. Consider this - an LED requires a voltage difference across its terminals to operate, and its brightness varies as the amount of current. Therefore, you are controlling the current with the change of voltage. Remember this formula - V = I * R, where V is voltage in Volts, I is current in amps, and R is resistance in ohms. It is also common to use milliamps and kilohms, which yield volts in either case, but are more convenient to use.

There are two basic setups. The first is very simple, and works good for what would be considered a power indicator. The circuit is as follows:

(+Voltage Source)---------/\/\/\/---------|>|-------(GROUND)

The resistor limits the current from the +12V source. The amount of voltage is only relevant to bias the junction and (ultimately) the power rating of the resistor. The amount of current is critical, however. Too much, and the device will burn out. So, let's start at a value of, say, 50mA. If V=IR, and we know V=12V and we want I=50mA. This gives us: R=V/I=12V/50mA=0.6kohm, or R=600ohms. The power rating on the LED depends on the current it consumes. Since P = I * V, where P is power in Watts, and I=50mA=0.05A and V=12V, P=IV=(0.05)*(12)=0.60W, so you would need around a 1/2 Watt rated resistor. Now, these are just SAMPLE numbers, and you have to experiment to see what current rating works for you. Note that I'm also ignoring the inherent on resistance of the LED, which would be added to the R of the circuit and would limit (a very tiny amount of) the current.

The other configuration is where you actually sink through something that can supply voltage. The configuration is like this:

(Voltage Source)---------/\/\/\/---------|>|---------(Signal Source)

In this configuration, when the signal source is at zero volts, current sinks through it and is limited by the resistor. When the signal source is active, the voltage drop across the resistor becomes low or nothing since both sides of the circuit are at the same potential. Use EXTREME caution if you connect anything in this way, since if you exceed the current sinking capacity of the signal source, you could burn out the driving device and do permanent damage. The principles are the same as the above with the exception that in V=IR, the voltage is the difference between the voltage source and signal source.



<<
The table on the back of the LED package was confusing Is there any way to make my hdd light have a blue light? I would have to plug it into the +2v connector from the mobo so it knows when to blink - but there isn't enough power to light the darn LED if I do that!
>>



There are probably two ways to do this. First, is there any way you can use a multimeter to measure the voltage that goes to the HDD LED (measured between the LED and the case)? If it's something around 5V, then you could probably put an LED there. IF IT ISN'T, you need to build a circuit that uses the original LED source as a trigger, and tap off your case power supply for it. Go find an application note for an LM386 op-amp (probably at http://www.natsemi.com, power it off the case power supply, put the input from the HDD LED source, and the output through a resistor to the blue LED. I don't know enough about your setup, so I can't say for sure what circuit is best, and there are a million variations on how to do it.



<<
For the resistors, on the back of the package (the LED package), it doesn't say what wattage to use, it just says 680ohm for 12v DC. I think I am missing something here. Any clarification for my cloudy mind would be greatly appreciated. >>



Well, if you do the power calculation above, we got to 600ohms 1/2 Watt rating. So just use 1/2 Watt.

One more last warning - if you exceed the current sourcing capability of the device driving the LED, you will get voltage drop in the device, which could affect operation and possibly damage the driving device. If you're not sure, measure the current and use the calculations above to see what you would get with the resistor in line. If in doubt, just build the circuit with the LM386 and you should be ok. Good luck, and don't burn anything out.

 

[Servanya]

excellent!
I think I may try building a circuit of LEDs to light my case up! It would be more fun than just buying a cold cathode tube from thinkgeek.com.

 

[CTho9305]

what is a "cold cathode"?

 

[Floydian]

i think its just a flourescent light, as in CCFL, cold cathode flourescent light - correct me if I'm wrong, I'm not totally sure. They come in various colors and stuff, apart from the usual white.

 

[Superdoopercooper]

<< hmm, that's a bit of pickle you got there What I would do is use the +2V to drive an NPN transistor which in turn drives your LED. That way you can use the +12V to power the LED, while using the +2V to drive it. If you're not into electronics you probably don't have a clue how to use a transistor. So here's a small schematic that will -hopefully- clarify it I'll leave the implementation of this thing in your case to you >>



Use a transistor to turn on a diode?

An LED generally (usally red/yellow ones... not sure about blue ones)... needs about 5-10mA to be fairly bright For a blue one, say 10mA. So, if you have an LED voltage of, say 4V, then you have 12V - 4V = 8V. Therefore, to get 10mA out of that... you get 8V/Resistance = 10mA... ==> 800ohm. So, the thing on the diode package makes complete sense IMO.

So, as long as your supply can supply # of diodes * 10mA... then you can wire as many of these 800ohm (680ohm) and blue LED combinations in parallel as you'd like.

Hope this ads clarity to the already fine answers given by other folks. BTW: dont' use a transistor to power a diode. That is just silly. No offense phalanx.

 

[JW310]

<< BTW: dont' use a transistor to power a diode. That is just silly. No offense phalanx. >>


Your idea would work for using a blue LED to show that the computer or a fan or whatever is on. However, how would one go about powering an LED that is controlled by the motherboard (such as the HDD activity LED)? If that were the case, then wouldn't the transistor work fine? Since the LED needs 4V and the motherboard would only supply 2V, without using something to step up the voltage, then I'd say that I don't see any reason why phalanx's suggestion of using a transistor to switch the LED on and off wouldn't work. Yes, using the transistor would be severe overkill for using the LED as just a power indicator LED, but he'd need something such as phalanx's circuit to use the 2V to power the LED.

JW

 

[Servanya]

Cold cathode tubes are these.

And...these LEDs are rated at 20mA. I was thinking about running a bunch of lights in parallel to light up the case - and then use the transistor for the hdd activity light.
Happy new year!

 

[Superdoopercooper]

<<

<< BTW: dont' use a transistor to power a diode. That is just silly. No offense phalanx. >>

Your idea would work for using a blue LED to show that the computer or a fan or whatever is on. However, how would one go about powering an LED that is controlled by the motherboard (such as the HDD activity LED)? If that were the case, then wouldn't the transistor work fine? Since the LED needs 4V and the motherboard would only supply 2V, without using something to step up the voltage, then I'd say that I don't see any reason why phalanx's suggestion of using a transistor to switch the LED on and off wouldn't work. Yes, using the transistor would be severe overkill for using the LED as just a power indicator LED, but he'd need something such as phalanx's circuit to use the 2V to power the LED. JW >>



Ahhhhh.... my bad... I stand corrected. I didn't really understand why he was saying to use a transistor to power the diode. So, you're saying the control voltage... is only 2V. But since you need 4V, you will have to make some sort of Transistor or op-amp buffer which is turned on by the 2V, but that is powered by the PC's 5V or 12V supplies.

Ok... that would be A-OK. Sorry about that. It just seemed like he was saying... "you only have 2V, so throw a transistor in... and youl'll have 4V" He made it sound like the transistor was POWERED by the 2V supply. That wouldn't work. hehehehehe..... Plus it was late and I probably mis-read as I was falling asleep at the desk.

Yeh... you'd want some sort of common collector transistor circuit to power the Blue LED. That would act as a buffer if I'm not mistaken (it's been 4 years since I last even looked at a transistor circuit).

Anyone know why blue LED's require so much voltage? Someone may have said it here... but I'm too lazy to re-read, so could someone give me a one liner?

HAPPY NEW YEAR!!! Hopefully 2002 will be GREAT!

 

[Servanya]

I couldn't find a reason for them requiring more voltage. It seems that most high-output ones require 4-5v. I found many websites that explain how they make blue and white light with phospors and stuff, but no info about voltage requirements (other than "some require more").

That's a good question.

 

[Servanya]

In case people are interested - I have decided to take on this little project:

Link here!

Looks even cooler than my idea! Now my fans can run quiet when I want to.

A simpler switch is here

 

[Delbert]

I have been dabbling a little in some case modding recently including some LED stuff. I still have some questions myself, but I did find some good LED information here and here.

view my LED challenge
here
and
here.

 

[Servanya]

Dang! That fan mod is a great idea. I may try it out too.
VERY cool.

 

[Delbert]

I?d better give credit where it is due. The info for the fan LED I found here.

I couldn't figure out what to do (resistors to use) when running two LED's in parallel. Then to complicate matters I wanted two different colors and couldn't find a red and blue with the same value (voltage/ma). I wound up winging it and running them in series. God watches over ignorant electron idiots I guess as it has been running for several weeks now no problem.

 

[malapala]

I am not very good at electronics, but I think it works like this:

[edit] Nope, that is not how it works... =(

 

[Cosmo56]

Hello there, I was just looking at the Schematic given by Phalanx, and it looks like something is up, wouldn't you have to use the -12v supply if you were planning on using the NPN transistor that he is talking about?

 

[phalanx]

erm, I don't see why you would ever need a -12V source to drive a LED with a transistor
The only thing that you could argue about is the base resistor, which seems a bit large after all because you only have 2 V to drive it.
So that would give 1.4 V over the resistor resulting in a base current of 140 µA.
Since the LED draws about 20 mA that would require an HFe of 142, but the transistor should be able to handle that.

 

[Hanpan]

<< Anyone know why blue LED's require so much voltage? Someone may have said it here... but I'm too lazy to re-read, so could someone give me a one liner? >>



I don't have a long techincal answer for you at the moments but in a nut shell.

Briefly, the LED chip consists of two different types of semiconductors that have been joined together. The LED chip at the junction of these materials emits photons of a specific wavelength in response to a flow of electrons. These electrons are transported upward across a band gap between the valence band and the conduction band by an external voltage V, called the cut-off voltage. This gives each electron a specific extra potential energy E= eV (e being the charge on an electron), equal to the energy width of the band gap. Once the electrons are in the conduction band they can flow from the one semiconductor material with more electrons the other material with less electrons. Then the electrons fall back down across the gap in the other semiconductor material to the valence band, emitting a photon of the same energy (E =hf). This emitted photon is the light we see from the LED.

Light's colour is determined by it's wavelength lamda. The frequency (f) in the the above equation (E =hf) is (speed/wavelength) or c/lamda. Now remeber the visible spectrum represented by ROYGBIV. Basically the farther left the colour of light the greater the wavelength. So red orange yellow and green photons have relatively long wavelengths about 650nm for red, while blue light has a wavelenght of about 435nm or so . Taking the earlier equation of f=c/lamda we see that as lamda decreases f increases. Since E varies directly by f (remeber h is constant) we see that light with shorter wavelength has as higher frequency and corresponding energy. Therefore in order to emite a photon of higher energy, more energy is needed.

Now please be aware that this is somewhat simplified as one cannot simply always add more enegery to get any colour ie (red led with loads of energy will not give you blue instead it will burn out) but this should be good enough to explain why blues leds need more voltage.

 

[AnthraX101]

I just slapped in a 5v blue LED from RS on my Asus A7A266 and it worked fine.

Armani

 

[Servanya]

Damn, I have heard of people just swapping them out before. That sucks - err, good for you Wish mine was that easy.

What is "RS"?

 

[Hanpan]

<< Damn, I have heard of people just swapping them out before. That sucks - err, good for you Wish mine was that easy.

What is "RS"? >>



Well in a nutshell

RS can described as a most peculier outfit. It is found in many malls and shopping arangments and even has an online site dedicated to it. RS can vary from place to place and no two RS are exactly alike. They even have RS in canada *gasp* though I do not believe they are available in Europe yet.
















RS= Radio Shack

 

[StandardCell]

Radio Shack also = Rat Shack

Actually they're good. I also forgot to include (like a moron) the voltage drop through the diode in my equations. Doh!