LEDs Brightest When New?

[jayflash]

Do I understand correctly that the forward voltage (Vf) of LEDs drops as they age? Wouldn't that infer greater output as the LEDs age? My experience has been the opposite for all types of 5mm and Luxeon powered lights that I own.

Without exception, all of my LED lights were brighter during the first hours of use. They all decreased roughly 10% - 20% in a short time and then leveled off, more or less maintaining their output. NewBie offered much info regarding the degradation of LEDs due to heat and over optimistic power ratings, but I'm talking about less than 10 - 20 hours of use.

My observations aren't accurate enough to be scientific and I may be in error regarding some of my lights, but overall the metered output indicates my observations are, mostly, correct. I use a few "standard" incandescent sources to periodically check my light measuring set up for stability. Could the solar cell I use be "aging" and losing sensitivity to blue light? My incans have not lost measured intensity.

Might the associated drive circuits all be at fault? Any ideas? Thanks for any thoughts and information.

 

[tebore]

Well the phosphor does start to loose it's effect as it ages. I notice that LEDs are not at their brightess when out of the box. It's right after the "burn-in" period where the Vf shifts/lowers and right before the phosphor starts to degrade. It's usually in the first day or 2 that this happens. This is visually speaking.

 

[2xTrinity]

My observations aren't accurate enough to be scientific and I may be in error regarding some of my lights, but overall the metered output indicates my observations are, mostly, correct. I use a few "standard" incandescent sources to periodically check my light measuring set up for stability. Could the solar cell I use be "aging" and losing sensitivity to blue light? My incans have not lost measured intensity.

That is an interesting possibility -- I know that many forms of plastic "yellow" as they age, that is, they filter out more and more blue -- that would affect the output from the LEDs but not so much from the incandescent.

If the losses are due to phosphor degradation, one idea might be to take two LEDs that have similar tint, one run for several hours, then compare -- if the tint has shifted, it's probably due to the phosphor on the LED degrading, rather than the LED itself degrading. I suspect that will be the most common mode of failure for white LEDs, not the actually emitter wearing down. (the green "num lock" LED on my keyboard has only very slightly dimmed from almost 5 years of continuous on time, doubt any sort of phosphor LED would hold up as long)

 

[Patriot]

This seems to be what I've experienced also, with my L2s and L4s. Whenever I purchase a new one it's noticably 10% - 20% brighter that the old one.

I've even nocticed this a PD1-CE that I was doing run time tests on. It now outputs about 90% of the one that hardly has any time on it. :thinking:

 

[MikeSalt]

Well, the CREE 7090 X-RE specification state that it maintains at least 70% output after 50,000 hours service. This implies that the output decays with age. But 30% over 50,000 hours is a pretty small gradient, lol.

 

[gammaray1965]

So much to be desired!

 

[2xTrinity]

Well, the CREE 7090 X-RE specification state that it maintains at least 70% output after 50,000 hours service. This implies that the output decays with age. But 30% over 50,000 hours is a pretty small gradient, lol.

Well, I suspect the degradation is exponential, rather than linear -- meaning it will degrade fastest near the beginning, and then less and less as it goes along. However, from the description in the original post, 10-20% in the first couple dozen hours is FAR more severe than that. I might expect that sort of degradation after several thousand hours, not 20 hours, based on the specs from Cree. It's possible that the driver that is pushing them isn't well regulated though, a high voltage spike or something might damage the LEDs.

 

[Flying Turtle]

I think I've noticed a dimming of my old first generation Peak Matterhorn 3LED. It gets used often with a diffuser as a kind of night light, so it's been racking up the hours. Hard to estimate how many, but probably close to a thousand by now. Just doesn't seem as bright on a freshly charged battery as it once was.

Geoff

 

[Calina]

Newbie, if you read this, we need your lights!

It could be on your own site...

 

[xiaowenzu]

I've even nocticed this a PD1-CE that I was doing run time tests on. It now outputs about 90% of the one that hardly has any time on it. :thinking:

I think this has probably to do with the lights being over-driven. YIKES!

 

[TORCH_BOY]

I have also noticed the same with the 5 mm leds

 

[cslinger]

This might be a problem for permenant use lighting applications like home lighting, car headlights etc. but for flashlights I think it is kind of a moot point. Most people who use flashlights as tools will never notice this over the lifetime use of their lights, would be my guess. Even a lux five rated between 250-500 hours before dimming is still looking at an average of around 400 or so battery changes and even at that point it is still likely putting out useful light.

LEDs are not perfect but I think they are a big step in the right direction. I will take dimming over time vs. the pop of a busted bulb when I am in the middle of using it anytime.