wiring LED's

[Aero]

Ok, I just wanted find the best way to do this before I try it. I want to wire in about 15-20 LED's for my mod. Woud the best way to do this be to just take off the molex connector and then just wire them all in parallel, and solder them togather? And should I worry about the amount of power that will take from the system?

[crazybillybob]

Aero,

As the molex only has 5Vdc and 12Vdc running on it and leds take from 1.5 to 3.5 Vdc, if you wire them up in parrell you'll get a nice flash when you power them on but they will not last long.

1st allways use a current limiting resistor with LEDS.
2nd The best way to wire them up (power usage) is series parrell (a few leds inseries wired parrell to another group in series.)

Check out this web site enter the voltage, number and voltage of LEDs, current rating and it gives you the best resistor and wireing diagram.

http://led.linear1.org/led.wiz

Also make sure the PSU can handle the extra load, overloading a stressed out PSU is an easy way to destablize your pc!


(most red and yellow leds are 1.5v blue and white are 3v.... Ultrabright and uv very ....but allways check with the supplier for correct specs on the leds your going to use.)


Good luck,
Crazybillybob

[DaveW]

Hey Aero, no worries about the LED's. They take up so little power that you can probably run about 150 off them before you really noticed a drain.

Wire your LED's in parallel and that way the voltage isn't divided; they will all get an equal amount. If you spliced your Molex for the 5volts lead, then all you need to do is work out the voltage and the current, then slam a resistor in there. Each LED will have a rating, and you'll have at the very least 2 of the following peices of information: Power in Watts, Current drain in Amperes, or Voltage in Volts.

Now, the equation for power is P=IV (Where I is the current and V the voltage). So you can work out the voltage of the LED and it's current rating.

Now, you have 20 LED's. So you're going to want to ADD together the current, because they're in parallel. This is your current drop. As they're in parallel, the voltage is the same as the voltage for one resistor. The internal resistance of the LED's is negligible and can be ignored (And plus, the equation is 1/Rt = 1/R1 + 1/R2...1/Rx and who wants to work that out?)

Anyway, now you're cooking with fire.

It's easy here: You're getting a 5v supply. You've worked out V(led) and I(led). Now your equation is;

R=(V(supply)-V(led))/I(led)

The value for R is the type of resistor you'll need. Slap that resistor in series with your parallel LEDs. And that should be you. I hope that clears things up and isn't too difficult to follow. If there's something you're unclear on, don't hesitate to ask.

-Dave

[DaveW]

Damn you billybob, ya beat me to it!

-Dave

[xmastree]

Do not wire them in parallel. Since an LED is a current device, rather than voltage, any slight difference in forward voltage drop will mean a large diference in current. It's better to connect them in series, in groups.
So, if they're 3V LEDs, three in series will give 9V so the resistor will need to drop 3V at the required current.

So... R=V/I What's the current? say, 20mA
V/I = 3/0.02 = 150
So 150 ohms is good for running three 3V LEDs in series from a 12V supply.

+ve ---===------>|------>|------>|----- -ve

I was thinking the other day, I might just make a howto explaining LEDs, since they do seem to cause some confusion.

[DaveW]

My apologies-i checked my books and xmas appears to be correct. However, trying to wire 20 of these will be problematic. It's too late for me to draw this out; i'd love for my 500th post to be one of spectacular relevance but i'm exhausted. Plus, i just typed all that above.

Although, to be fair, mathematically speaking everything above is correct in an ideal system with no voltage fluxuations. Here's to the big 500 guys!

-Dave

[xmastree]

I'll knock up a quick drawing, I've just got up and the coffee is working its way round my system.

Reckon I should also work on an 'idiot's guide to LEDs'? since to me they're dead simple, but to some theyre very confusing. and I know why...

Here's one I made earlier: (get down, shep!)

http://i50.photobucket.com/albums/f341/chrispollard/LED.png

Which I did for tybrenis. It shows two methods of wiring three LEDs to run from a battery. Same principle applies. The second one is much better, since the total current drain for three is the same as for one. In other words, you get two of them for free.

[xmastree]

Here you go (Assuming 3V LEDs and 20mA, the resistor will be 150 ohms).
Each branch will draw 20mA, making a total of 120mA. An extra load of 1.44W, I think your PSU can handle that.
http://i50.photobucket.com/albums/f341/chrispollard/LEDs.png
If you're using lower voltage LEDs, you can fit more in series, reducing the current rain even further.

What's the specification (forward curent, forward volt-drop) of the LEDs?

[Aero]

lol, I really just wanted to know if wireing it directly to the PSU would work and make sure my already strained PSU wouldn't be taking too much of a load. And I already know about wireing LED's with resistors. Also it is better to use them in parallel with one resistor to each LED. Its more work but in the long run its much easier if you want to add more to the circuit, plus LED's are voltage based, meaning less voltage = less bright so series isn't a good choice. (unless you do it like xmastree, i think that should work:)) But thanks for the replies guys.


Also DaveW welcome to the 500 club, we have cookies.

[crazybillybob]

Also DaveW welcome to the 500 club, we have cookies.


Yeah, the 500 club has cookies but the 449 club has ........R2 :)
And less Hot Air!



Couldn't help myself......



Crazybillybob

[DaveW]

Hey, 400 of my posts were nonsense anyway.

Xmastree, you have quite a few resistors there, but you can manage with a single resistor in series unless i've forgotten something important.

Just for a sort of footnote.

Oh, a CBB, i have too much respect to get offended by anything you could say. Anyway, you make every post count :)

-Dave

[Aero]

I'm not sure if that would work though, becase while each series is geting teh same voltage, each LED in the series is getting progressivly less so they would just be dimmer in order wouldn't they?

[DaveW]

No, circuit's don't work like that. Voltage is potential difference; it's not like fuel that gets used up at each stage. It's more like rain falling into different sized buckets.

I know it's a crap analogy, but hey. I'm tired, it's the best i can come up with, and water analogies in electronics are more common than STDs in the Red light district.

-Dave

[Aero]

No, circuit's don't work like that. Voltage is potential difference; it's not like fuel that gets used up at each stage. It's more like rain falling into different sized buckets.

I know it's a crap analogy, but hey. I'm tired, it's the best i can come up with, and water analogies in electronics are more common than STDs in the Red light district.

-Dave

hmmm....well I'm pretty sure I understand how the work, I must be saying something wrong. (electricity was the only test I got an "A" on in Physics this year so I must understand something somewhere). bah, my brains too fried to thinka bout this right now, ACT tomarrow so sleep now.

[crazybillybob]

What daves trying to say is that you can limit the current in an entire branch with 1 resistor (this only works if all the LEDS are the same kind and color)


So here's a little drawing to help

................Res
+---------^^^--------|<-----gnd
..........................|----|<----Gnd



OK Ignore the ....... It's just so everything stay where I drew it!

Hope that helps.

Crazybillybob

[CanaBalistic]

I had it at the diagram but the more you guys talk the more i get confused. :)

[Omega]

I had it at the diagram but the more you guys talk the more i get confused. :)


Diagrams > Geeks rambling on

[xmastree]

I know it's a crap analogy, but hey. I'm tired, it's the best i can come up with, and water analogies in electronics are more common than STDs in the Red light district.Damn, what a time for me to lose my conenction... I've been chomping at the bit waiting to answer this.

Think of an LED forward voltage as an overflow in water terms. The voltage can never get higher than the level of the overflow. Try to raise it and it just flows faster. Feed it through a small pipe (resistor) and you have flow control.

I've prepared a simple explanation, but I've put it in its own thread, here (http://www.thebestcasescenario.com/forum/showthread.php?t=2669).

[xmastree]

I'm not sure if that would work though, becase while each series is geting teh same voltage, each LED in the series is getting progressivly less so they would just be dimmer in order wouldn't they?
No, the voltage across each would be the same.
If you measure from each one to ground, yes, each will be less. But if you measure across each LED they'll be the same.

As for only one resistor and LEDs directly in parallel. Not a god idea. If manufacturing tolerances cause one LED's voltage to be lower than the other, that one will take the lion's share of the current.

[DaveW]

Bah, who gives a crap anymore. Now i'm going to have to breadboard the simplest LED circuits, all because of this thread. Thanks a lot guys!

-Dave (who is only kidding ;) )

[xmastree]

Bah, who gives a crap anymore. Now i'm going to have to breadboard the simplest LED circuits, all because of this thread. Thanks a lot guys!

-Dave (who is only kidding ;) )

Got a variable power supply? Make a graph of V/I for an LED, Start from zero. Once it starts to light, go up in 100mV steps until you burn it out. :redface:

http://www.technology.niagarac.on.ca/courses/phtn1333/images/LEDI-VCurve.jpg (http://www.technology.niagarac.on.ca/courses/phtn1333/Lab2.html)

See how sharply the current rises once it starts?

If you have the means, measuring the light output would also be useful info.