confused about LED wiring...

[blk03MitsuES]

I already read a couple of times eclecticos wiring tutorial. but i'm still a little confused.

i'm interested in these leds (http://www.ledshoppe.com/Product/led/LE1004.htm) from ledshoppe.com.

Forward Voltage (V) : 2.8 ~ 4.0
Max Continuous Forward Current : 30mA

i figured i could wired up 4(3 x 4v = 12v from molex) so i tried the online led wiz just to see if i was right. and nope, i was wrong.
it's showing me that i need to connect two leds and 150ohm resistor twice.

so then i drop the voltage to 3.3v because thats an average for UV leds and then it shows i need 220ohm resistors...

is the wizard playing it safe by adding the resistors or am i ok using 4 leds without one?

2nd question, is cat5 wire(individual strands of it) good enough to wire leds or do i need a ticker/bigger cable?

3rd question, how many leds can i wire to an individual molex plug? or does it depend on my PSU?


btw i'm using http://led.linear1.org/led.wiz if there's a better one let me know. thanks

[mtekk]

For wiring up diodes in series (one after another), you can't use the max forward voltage as your assumed drop. The actual drop across the diode is something between .7v and the minimum number they provide. The best way to calculate what resistor you need (I'd always put one in) is to use ohm's law:
V = I*R

For 12v your V = 12v, now we can adjust this by subtracting the forward activation voltage off of 12v. Assuming a very conservative 1v, you could place 12 diodes is series, or do something reasonable and use 4, that would drop your voltage to 8v. Thus ohm's law becomes:
8v = I * R

Now for I we will set that equal to the Imax of the diode, in this case 30mA which is .03A. Thus the equality becomes:

8v = .03A * R

Solving for R yields 267 Ohms of resistance.

Applying the same thing, for the minimum forward voltage of 2.8v and 4 LED's in series yields:

.8v = .03A * R
R = 26.6 Ohms.

I would not use the 2.8v forward turn on voltage estimate, and instead I'd use the 1v estimate and place in a ~267 Ohm resistor.

CAT5 is definitely thick enough, we use something like it in size in EE Lab with LEDs so it will definitely work, though I prefer using stranded cable when possible.

LEDs do not draw a lot of power (P = I*V) (.12W at max specs) so you should be able to place at least 20 of them per line (2.4W per molex cable) which is still less than what other things pull from it.

[blk03MitsuES]

267ohm resistor for 4 leds or is that every 2 leds?

[mtekk]

That *should* be enough for four LEDs in series. The only thing extra resistance past the minimum will do is cause the LED to not glow as brightly as it could. Then again it is always better to be safe than sorry, so putting in 300 Ohms may be a better option, especially since I think you can get resistors in 300 Ohm sizes, and I'm not exactly sure is 267 Ohms is a standard size or not.

[blk03MitsuES]

270 ohm and 300 ohm resistors are just as popular on ebay. as far as the last color code do i go with silver, gold or none?

[mtekk]

Gold is a rating of +/- 5% (so a 300Ohm could be anywhere between 285 or 315)
Silver is a rating of +/- 10% (so a 300Ohm could be anywhere between 270 and 330)
No band is a rating of +/- 20% (don't even bother looking at these).

I prefer to stick with gold rating ones as they are pretty cheap but have higher tolerances than silver. Anything worse than Silver banded is not worth even considering as the swing in resistance becomes way too large.