Has any here hit the 50,000 hour led life span?

[ledmitter]

How long is 50,000 hours?

There are 8,760 hours in a year, so here's what 50,000 hours works out to be, depending on lamp hours of operation.

Hours of Operation:

24 hours a day 5.7 years
18 hours per day 7.6 years
12 hours per day 11.4 years
8 hours per day 17.1 years

Why is the life span of an LED measured as lumen depreciation?

The life span of an LED is vastly longer than that of incandescent, fluorescent or HID lamp sources, generally lasting 50,000 hours or longer. Although the LED never really burns out, product life span is measured by lumen depreciation.

The Illuminating Engineering Society's (IES) current standard for calculating the life of an LED as the point at which the LED reaches 30 percent lumen depreciation.

Remember, a 100,000-hour rating is not equivalent to lamp life rating. LED life is rated where it has reached 30 percent lumen depreciation. At 100,000 hours an LED would still be operating, but at a decreased lumen output.

What is junction temperature?

Junction temperature is the temperature at the point where an individual diode connects to its base. Maintaining a low junction temperature increases output and slows LED lumen depreciation. Maintaining a low junction temperature is critical for evaluating an LED product's quality and ability to deliver long life.

What can affect the LED junction temperature?

There are three things which can affect the junction temperature of LED's: drive current, thermal path and ambient temperature. In general, the higher the drive current, the greater the heat generated at the junction. Heat must be moved away from the junction in order to maintain expected light output, lifespan and color. The amount of heat that can be removed depends upon the ambient temperature and the design of the thermal path from the junction to the surroundings. Typically the junction temperature should be maintained below 120°C.

 

[ledmitter]

Lumen depreciation in LEDs varies depending on package and system design. The primary cause of lumen depreciation is heat generated at the LED junction. LEDs do not emit heat as infrared radiation (IR) like other light sources, so the heat must be removed from the device by conduction or convection. If the LED system design has inadequate heat sinking or other means of removing the heat, the device temperature will rise, resulting in lower light output. Clouding of the epoxy encapsulant used to cover some LED chips also results in decreased lumens making it out of the device. Newer high-power LED devices use silicone as an encapsulant, which prevents this problem. LEDs continue to operate even after their light output has decreased to very low levels. This becomes the important factor in determining the effective useful life of the LED.

 

[yifu]

I wonder what the CREE XML's lumen deterioration rating is after just 1,000 hours. These LED's are being driven pretty hard in many of these lights with questionable heat sinking compared to what they using in the lab.

That's easy, the lumen maintainence results are here http://www.cree.com/~/media/Files/C...Lamp/XLamp Application Notes/LM80_Results.pdf After 6000 hours of driving an XML at 2A and at 85 degrees celsius, the LED maintained 97.2% of its output. Not bad.

And one more thing, unlike other light sources with catastrophic end of life events like incans (burn out), mercury arc (unsustainable loss of arc gap length), LEDs will not blow out after 50 000 hours, but rather loss 30% of initial output. This extrapolated lifetime would is usually done at half the maximum drive current (750mA for XPGS, 1.5A for XMLs) so the L70 lifetime would increase if the LED is driven less, or decrease if overdriven.

 

[StarHalo]

the LED maintained 97.2% of its output.

That's not nearly enough to tell the difference by the eye in a direct side-by-side comparison with a brand new copy - insanely good; not only can you pass your flashlight to your grandkids, but it'll still be just as bright..

 

[ledmitter]

That's easy, the lumen maintainence results are here http://www.cree.com/~/media/Files/Cree/LED%20Components%20and%20Modules/XLamp/XLamp%20Application%20Notes/LM80_Results.pdf After 6000 hours of driving an XML at 2A and at 85 degrees celsius, the LED maintained 97.2% of its output. Not bad.

And one more thing, unlike other light sources with catastrophic end of life events like incans (burn out), mercury arc (total loss of arcs), LEDs will not blow out after 50 000 hours, but rather loss 30% of initial output. This extrapolated lifetime would is usually done at half the maximum drive current (750mA for XPGS, 1.5A for XMLs) so the L70 lifetime would increase if the LED is driven less, or decrease if overdriven.

Nice catch. Thanks for that. Super interesting reading.

An XML being driven at 2A would be approx how many lumens? Any conversion numbers come to mind Yifu?

I've always thought the 700+ lumen XML's where being driven upwards of 3A. Guess i'm wrong.

 

[yifu]

Nice catch. Thanks for that. Super interesting reading.

An XML being driven at 2A would be approx how many lumens? Any conversion numbers come to mind Yifu?

I've always thought the 700+ lumen XML's where being driven upwards of 3A. Guess i'm wrong.

Sure no problem, from Cree's lab results here http://pct.cree.com/ A XML of the U2 bin at 2A produces a typical 741.6 lumens, with a 7% variance either way. So the OTF lumens should be at least 500 lumens. The 700+ lumens XMLs are normally driven at past 2.8A, although the XML LED can be expected to output 1200 LED lumens at 4.2A, not very efficient at that current though.

That's the reason i find it funny when people try to avoid running their lights on high for fear of loss of output. The loss after a heck of a long time is really, negligible, provided you've got a reasonable heatsinking solution, which most good lights have anyway.

The tests on overdriving XMLs are here. The heatsink used is not very good, so the outputs could be potentially 200 lumens higher, as verified by some CPF members. It's the T6 bin BTW.

 

[luceat lux vestra]

Hold that thought............... I get back to you in about 5.7 years......................................................................................................

:devil::devil:

{sorry}

 

[Glock27]

My SC60w has now been "On" for 443 days. Mostly on Low1, but hundreds of hours on High. 10,600+ hours. Still can't tell difference on white wall between others in same batch.

G27

 

[ShineOnYouCrazyDiamond]

Not sure if it counts but my wife has a Sony clock radio with red LED light bars making up the numbers. She has had this light for close to twenty years and short of maybe times during a move it has been plugged in the whole time. The LEDs appear pretty dim to me at times but they are still running.

LEDs are semiconductors. Their true failure rate should be no more or less than any other semiconductor diode, transistor or IC circuit. People have LED power indicators on items all throughout their homes and they just keep working year after year after year.

 

[Echo63]

Some of the older hands here might recall reports of the LuxV exhibiting a visible decrease in brightness after as little as 500 hours of use. Apparently those were driven hard enough to get HOT in some lights.

Edit: And if anyone expects 50,000 hours of runtime from LEDs at 'CPF' drive levels, I have a 500,000 hours MTBF (i.e. 57 years) hard drive to sell them.

I had a lux V in an L4 die after maybe 400 hours of use
It got pretty hot though, and i think one of the screws holding the heatsink to the head was loose - which halved the transfer capacity of the thermal path from the led to the outside world (maybe more)

I think a mehanical failure is more likely than a led that has reached the end of its life
(i believe my failure in the l4 was heat related caused by a mechanical failure)
And i dont think Leds will reach their rated lifespan at CPF type drive levels - maybe in things like the LL malkoffs

 

[Jash]

I've noticed that once my lights turn around 18 months they get gifted to non-flahsaholics and something new comes in the mail.

Don't ever plan on reaching more than a couple of hundred hours with any given light.

 

[LEDninja]

I had an LED light bulb die on me after 1 year 10 months 24/7. That is 15,000 hours. I suspect capacitor failure instead of the LED though.

 

[Kestrel]

The LED dimming behavior most often observed here is that our current lights suddenly get dimmer when the 'new one' arrives in the mail, lol.

 

[rickypanecatyl]

So many led lights I have none last as long as an incandecent mag light. Something always breaks - probably not the LED but your pretty lucky to get 200 hours out of most LED lights you don't lose.

HID's are much worse. In the off road motorcycle world many companies selling HID's talk about how on rough, dirt roads HID's are so much more reliable as there is no filament to break. So many of those lights break in a few miles, most of them can't stand up to any level of abuse. But of course its something else that broke - say the ballast.

My point, it's just stupid to say (unless your a deceptive sales person) a LED or HID will last X times longer when some vital part of the equation will only last 1/2 as long at best.

 

[Illum]

It seems unlikely. 2000 consecutive 24-hour days of use, or 5.7 years solid. At rated current. Most people are not interested in such tests. And that isn't the life-span. LEDs are rated to 70% brightness loss. So at (xthousand) hours the LED is 30% dimmer.

The farthest I have gone was a string of CREE XREs in an semioutdoor setting and ran for almost three years 24/7... They have died as a resultant of exposure and corrosion as opposed to other things, it is difficult to know what decrease in intensity acually occurred and what decreases are from the deterioration of the jelly dome

 

[HighlanderNorth]

Id like to see if reaching that number before you buy another light due to better technology or what ever the case may be is likely.

I'd be totally shocked if I found someone who has put 50,000 hours on their flashlight........before losing it! 50,000 hours is a LONG time! Thats about 6 straight years of constant on........365 days per year, 24 hours a day.

Even if you only used the light for 4 hours per night, and had it on constantly for those 4 hours, for 365 days per year, it would still take you 36 years to hit 50,000 hours!

 

[JWP_EE]

I have a CREE P4 Star mounted on a 5" X 5" aluminum plate running at 2W used to light my computer area. It has about 30,000 hours on it and still going strong.

 

[bbb74]

Are electrolytic capacitors used in most of these regulated led lights?

They seem to be rated up to about 2000 hours of life. As they age their ESR goes up too.

Wouldn't this limit our lights to a much lower lifespan?

 

[KENNY LEE]

It's too long...

 

[HighlanderNorth]

I have a CREE P4 Star mounted on a 5" X 5" aluminum plate running at 2W used to light my computer area. It has about 30,000 hours on it and still going strong.

When did that P4 Star first come out? Does it give good brightness at 2W?