SparkFun Forums

[reklipz]

OK, lets say I have a 3.3Volt LED, and a 3.3V power source (the 3.3 is exact for this LED...), I would then not need a resistor inline with the LED?

Oh well, that isnt very likely to happen...

[Philba]

false. you will still need to limit current either by a resistor or a constant current circuit. the LED has no inherent current limiting ability.

I've often wondered how this works. It's a bit non-intuitive when Vcc == Vf. How do you calculate the amount of "drop" to then calculate the resistor value?

Anyway, I played with SPICE and found that the effective Vf seems to drop proportionally to the size of the resistor. the problem becomes one of accuracy. accuracy of the supply, current limiting resistor value and the actual Vf of the LED. I = V/R. for very small values of R, any error will be magnified and thus the current will either be too low (dim LED) or too high (very bright for a few moments then very very dim...).

[reklipz]

...
for very small values of R, any error will be magnified and thus the current will either be too low (dim LED) or too high (very bright for a few moments then very very dim...).

That gave me good laugh

OK, I think we've almost beaten this topic to death, so I guess I'll lay it to rest for the time being.

Thanks everyone for the awesome responses! (155+ views, 2 pages in 24 hours, not bad!)

[oPossum]

OK, lets say I have a 3.3Volt LED, and a 3.3V power source (the 3.3 is exact for this LED...), I would then not need a resistor inline with the LED?

Oh well, that isnt very likely to happen...

Yes, you can directly drive it with a voltage source. Look at the V/I curves in the spec sheet - they show how much current it will draw at a specific voltage. It is very important that Vmax is not exceeded - even breifly.

An easy way to get the exact voltage you need from a ~ 12 V source is by using a Point-of-load converter. These are used to reduce a 12 volt supply down to the 1.8, 2.5 or 3.3V used by high speed current hungry components. They are typically 85+ % efficient and quite small.

For $11 you can get one with an adjustable 1 to 5 V output at up to 6 A - enough for 200+ LEDs.

link

[bertrik]

OK, lets say I have a 3.3Volt LED, and a 3.3V power source (the 3.3 is exact for this LED...), I would then not need a resistor inline with the LED?

Oh well, that isnt very likely to happen...

Yes, you can directly drive it with a voltage source. Look at the V/I curves in the spec sheet - they show how much current it will draw at a specific voltage. It is very important that Vmax is not exceeded - even breifly.

Those curves can drift with time and temperature.

Here's an example for a Luxeon LED:
http://candlepowerforums.com/vb/showpos ... stcount=88

Say you used a new LED and set the voltage to 4.1 volts to get 1100 mA current. After 30 days of operation, Vf has dropped by 0.2V and that same LED would be drawing 1700 mA at 4.1 volts, an increase of more than than 50%.

[seulater]

http://led.linear1.org/1led.wiz
&

http://members.misty.com/don/ledd.html

[Bushman]

Waaaaaaaaait a second.

I thought that the 'forward voltage' rating is the voltage required for the LED to begin conducting and outputting light.

So for a LED with a 2V 'forward voltage', at 1.9V you get no light, and at 2.1V you start to get light.

The min/max description for forward voltage describes what the forward voltage values will be at min or max based on their design specs.

So if it says 'foward voltage max of 2V' they don't mean you can only run these things at 2Volts max. They mean "in general the max forward voltage required for our LEDs is 2V, but it might be less."
For min foward voltage they mean "in general the least amount forward voltage required for our LEDs is whateverV, but it might be more."

Am I completely wrong here?

I've never had any trouble whatsoever running LED's at 12 V. Just always remember to limit them to 20ma max and you are ok. (In fact you only need about 5ma generally unless you are trying to squeeze lots of light out of them)

[transcendentnb2]

I thought that the 'forward voltage' rating is the voltage required for the LED to begin conducting and outputting light.

Correct. Usually they give a min, typical, and max voltage value. This varies depending on the current being driven through the LED.

Say your LED has a min, typ, max of 1.7, 2.0, and 2.4v respectively. 1.7v could possibly do it, but you'll get very little current or light. This all depends, though, on the quality of the LED and it's manufacturing conditions (semiconductors can be moody). The more current that is driven through the LED, the larger the forward voltage drop will be.

Two important little charts are the Forward Voltage vs. Forward Current curve, and the Luminous Intensity vs. Forward Current curve. You can get some light being emitted at the minimum voltage, but it'll be very small.

Remember, LEDs are tricky things to fully understand (that goes with all semiconductor devices). The driver for the luminousity is the current, but you have to have a sufficient voltage applied across it to bridge the forbidden energy gap and get the ball rolling, so to speak.

But, as a rule of thumb, just put a current limiting resistor in series such that your desired current = (Vcc - Vled) / R. Hence, your Vcc should be greater than Vled. *Not* putting in a resistor is poor design, in which case your current will be limited by the natural resistence of the semiconductor and the connecting paths... which is tiny, hard to calculate, and not something that should be relied on for this application.

P.S. If you really want to get crazy, google around to see how LEDs can actually act as light sensing devices!

[Bushman]

P.S. If you really want to get crazy, google around to see how LEDs can actually act as light sensing devices!

Yeah. sweet.

http://hackedgadgets.com/2006/04/14/led-touch-sensor/

[reklipz]

... you have to have a sufficient voltage applied across it to bridge the forbidden energy gap and get the ball rolling, so to speak ...

Which is the same with all diodes, about .7V for a silicon diode, and .3V for a germanium diode.

Ok, this is really starting to make a bit more sense.

One bad thing about buying from ebay... no actual datasheets...

[seulater]

http://www.csgnetwork.com/ledsingresistancecalc.html

&

http://led.linear1.org/1led.wiz

&

http://members.misty.com/don/ledd.html

[reklipz]

http://www.leds.de/shop_content.php?coID=14

[Philba]

sigh. posted with out lookin to the end of the thread. my bad.