LS vs 5mm -Continuous- Runtime Limits?

If a light is to be used in a manner where it will run for long periods of time continuously, rather than briefly on/off, is there a concern with heat buildup in the Luxeon based lights? Advertised runtimes are an estimation of total life on a set of batteries, not a thermal limit.

I realize that the answer will differ depending on the design and heatsinking of the lights, but I know that both the 1 and 5 watt LS can produce significant amounts of heat if left to run constantly. Does this present a problem in terms of reliability or life? If so, are light with multiple small LEDs better for constant illumination?

No problem with running LS at their rated power level for extended period. If the light has reasonable means to conduct the heat away from the emitter, it's not going to over-heat. When you look into over-driven lights, that becomes a different issue.

[ QUOTE ]
LED mods As Small As Possible said:
When you look into over-driven lights, that becomes a different issue.

[/ QUOTE ]

Tell me more!

In every electronic device, there is always a percentage of loss in conversion - this is converted to heat as "waste"

If this heat is not able to be dissipated or pulled away from the device, it will damage the device

The Luxeon has a metal "slug" on the back of the package which allows easy conduction of excess heat from the die to an external heatsink

The typical 3mm and 5mm LED package, only one of the lead frame holding the "cup" is conducting heat out of the package; this very very small leadframe/wire is unable to properly conduct any excess heat out of the package

Some companies put the LED into a large metal case expecting that to couple the heat from the LED die to the case, but this is useless since epoxy is a very poor conductor of heat

Over current/heating on a LED could cause:

) Boiling of the Phosphor

Excess heat will start boiling or destroying the phosphor coating on the LED; this will cause the LED to start losing its White and become more and more "bluer"

A White LED is actually a BLUE LED covered with a special phosphor that converts the short wave approx 400nm light to a combination of light - ie White; this is much in the same way that a FL light works from exciting Mercury vapor at 245nm and 365nm

) Die destruction

The actual LED die will become damaged due to the heat and fail

) Bond wire failure

Excess current could cause the bond-wire that connects the lead-frame of the LED to the actual die to come disconnected

I have seen some cheap keychain lights actually blink caused by the bond-wire becoming disconnected from the LED die and then cooling down and reconnecting

) Microfracture -> Contamination

Excess current could cause the die to overheat and have microfractures/cracks of the epoxy resin, this will/could allow outside "air/etc" to enter the package and contanimate the die

The companies that make the LEDs clearly have designed them for a specified current/temp range; exceeding those will cause shorter operating life

Thanks for the write-up Inretech. Interesting stuff!

as far as over-driven lights, if they are in large bodied light that can keep it cool, then not a big deal. but if you're talking about say a super small light over-driven, meant only for short runs of illumination, and not to be sat down somewhere while running, then you can run into problems.. ..

[ QUOTE ]
INRETECH said:
...A White LED is actually a BLUE LED covered with a special phosphor that converts the short wave approx 400nm light to a combination of light - ie White...

[/ QUOTE ]
Let me make an itty bitty correction here if I may, for the benefit of all who read your message...the blue wavelength of the die in a typical white LED is 450-460nm, not 400nm. /ubbthreads/images/graemlins/wink.gif

Yes, you are correct - my mistake

Here is a plot that I took of a Luxeon, you can see the Primary LED

http://www.inretech.com/pictures/4.jpg

Not to stray too far off-topic... but do you happen to have the plot of a warm white? I am assuming the wavelength plot should be more evenly spread, especially with a CRI of 90.

I made these plots myself when I borrowed a Spectrometer; unfort - I had to give it back; but the owner will run graphs for me if I ask him nicely

Hmm yeah it would be interesting to see a graph of a warm-white. I bet a lot of people on here would want to see that info.