Can running an LED on too low power cause damage to it?

I haven't posted in some time, but I have a question.

Can running an LED on too low of power for a given amount of time cause damage to the LED, similar to running an LED at too high of power?

I'd be tempted to just say "no"

But I'll say that no, not that I've ever heard of or seen in my own messing around. If that were the case then current or voltage limiting as a way of dimming them wouldn't work as well as it does.

I too say no, as an LED is nothing more than a diode (...) and if you don't have enough voltage, light just wont be emmited. Unlike an incandescant bulb, running on a lower voltage will not increase lifetime. Overdriving an LED is another story, however, they still usually last a long time.

No

D-Dog, if it is a white led you will propably extend its lifetime due to that it run cooler and thus reduce the degradation of the phosphor...

cool! Thanks for the quick response's as usual!

D-Dog said:

I too say no, as an LED is nothing more than a diode (...) and if you don't have enough voltage, light just wont be emmited. Unlike an incandescant bulb, running on a lower voltage will not increase lifetime. Overdriving an LED is another story, however, they still usually last a long time.

Click to expand...

D-Dog is correct! :thumbsup: A very good explanation. I would point out though, that with an incandescent, there is a point where too low a voltage actually decreases bulb life. If the filament is not run at a high enough temperature, the vaporized tungsten will not redeposit itself back onto the filament. When this has happened, you will notice a grey/black film on the inside of the envelope. This tungsten is displaced from the filament, making it thinner and eventualy it will break, usually with a bright flash.

Dave

As I understand it, there are two basic ways to lower the output from an LED: lower the voltage, or supply square wave voltage at a relatively high voltage to the LED. The latter is known as PWM or Pulse Width Modulation. (I hope that explanation is clear).

Anyway, my question is this: Does PWM stress the LED more than simply running the light at a constant, lower voltage? A real world example would be to drive your car more slowly by flooring it for half a second and then letting up and allowing the car to coast for ten seconds, then repeating.

PhotonBoy said:

Anyway, my question is this: Does PWM stress the LED more than simply running the light at a constant, lower voltage? A real world example would be to drive your car more slowly by flooring it for half a second and then letting up and allowing the car to coast for ten seconds, then repeating.

Click to expand...

PWM does not stress the LED any more than it would if the LED were driven at the peak current. Your car analogy is more or less correct however, it'd be more like flooring and coasting at, for example, 200 times per second thus, smoothing it out so, you wouldn't really notice it!

Dave

45/70 said:

PWM does not stress the LED any more than it would if the LED were driven at the peak current. Your car analogy is more or less correct however, it'd be more like flooring and coasting at, for example, 200 times per second thus, smoothing it out so, you wouldn't really notice it!

Dave

Click to expand...

Actually, you can damage an LED with high peak PWM currents, even though the average current is well under specifications. As long as the peak current is within specs, there should be no problems.

Hi there,

There is nothing wrong with running an LED under the current spec except
for color output. Most of the LED colors are spec'd at their nominal current.
This is necessary in order to produce the correct color output.
Usually it doesnt matter that much because the color doesnt change
that much with reduced output. I've had no problems with this ever.
There's never been a problem with flashlights that i've heard about.

There are other types of applications that are more sensitive to the
wavelength of light however...

In dental applications it is possible that the dental epoxy used
for tooth repair may not cure properly if the wavelength of the emitted LED
light is not correct (special blue LEDs). For this kind of application it is mandatory to
supply the LED with the correct current level.

Nope,

will running on low voltage and super high current damage LED?

lets say luxeon I white, 0.05V,4A

chesterqw said:

will running on low voltage and super high current damage LED?

lets say luxeon I white, 0.05V,4A

Click to expand...

It simply cannot be done!

PhotonBoy said:

As I understand it, there are two basic ways to lower the output from an LED: lower the voltage, or supply square wave voltage at a relatively high voltage to the LED. The latter is known as PWM or Pulse Width Modulation. (I hope that explanation is clear).

Anyway, my question is this: Does PWM stress the LED more than simply running the light at a constant, lower voltage? A real world example would be to drive your car more slowly by flooring it for half a second and then letting up and allowing the car to coast for ten seconds, then repeating.

Click to expand...

Depends on the duty cycle and duration of on-cycle.
LED should be able to take 400Hz 100mA @ 25% duty cycle as if it was continuously 25mA, but if you were to drop it to say 1/4 Hz 100mA @25%, 1 second on, 3seconds average current over a long time is the same, but the on-time at 100mA exceeds the acceptable on-time duration.

chesterqw said:

will running on low voltage and super high current damage LED?

lets say luxeon I white, 0.05V,4A

Click to expand...

You can't push 4 amps through a LED with only 0.05 volts of pressure.

This is like asking if you can damage a 1 inch water pipe with 0.05 pounds per square inch of pressure with 4,000 gallons per minute flowing through it.

Toshi

Nope - no damage potential. BTW I've got an LED in a garage door switch that's been flashing on and off about 2-4 times per second for many years - it runs this way 24/7 to indicate the door is electronically locked. It's not showing any sign of dimming or quitting (it is probably just a 5mm LED).

meep. my physics suck

how about 2.5V and 4a?

It still can't be done. The LED must operate on its I/V curve.

Read these and be enlightened:

http://candlepowerforums.com/vb/showthread.php?t=77221
http://candlepowerforums.com/vb/showthread.php?t=72528

Hi there,

You have to have enough voltage AND current to run an LED in the first place.
The voltage is not proportional to the current, but as the current rises so
does the voltage. In other words, if you have a 3.3v LED at 350ma to get four
amps through it (not recommended) you would have to raise the voltage above
3.3v.

Too low won't hurt it, and most LEDs are rated for "pulse" modulation for higher output situations... for instance a Cree XR-E is rated (from memory, correct me if I'm wrong) for 1.8amp operation at 10% of a 1khz cycle.
If you want to dim them, 2 chioces are "lower power" or "pulse" modulation...
Pros/cons...
Lower power -
Pros: Simple (stick in a resistor/take out some current, etc...), cheap, LEDs are more efficient at lower powers (to their Vf cutoff point).

Cons : Wasteful... a 20mA rated LED (or a larger LED/array for better comparison), at 10mA uses 1/2 the power of it's rated output. (Sounds good, but see below)...

"Pulse" modulated LED
Pros : Extremely efficient,
--if a 1000khz frequency at 10% duty cycle is used, at rated 20mA current, the LED (stationary) will "look" the same to the eye as a full 20mA LED, but using only 1/10th the power. Eyes cannot easily detect frequencies above around 100hz from a stationary source. If source is moving, "flicker" can be detected, frequency and duty cycle can be tweaked further to make this undetectable.
-- Using lower power input, (10mA), combined with a "visually stable" pulse (say 500hz/20% duty cycle), the LED would appear roughly identical to a 10mA constant, but using only 1/5 the power (or 2.0mA effective).
-- Using 20mA input, 1000khz, at a smaller duty cycle (5%), the eye might see a drop in output... tweak as desired for "50% dim", etc). Your lower limit is the speed of the LED itself (Off-On-Off cycle time, typically in the 10 nanosecond to 1 microsecond range)
Cons:
Cost of circuits - ($$)
Cost of circuits - power to run the circuit
Space required for circuits
Tweaking to eliminate "flicker"