One of the great advantages of the LED is its reliability and long operating life. The light output of an LED tends to gradually decrease over the extended period of time, compared to a halogen or incandescent light source which fails abruptly. And the accepted norm for LED lifetime is L70, which means at the point when the light source is at 70% of its initial output. So when you see a 50,000 hour lifespan LED, it would imply 5.7 years if the light is operated for 24 hours in a day.
With LED technology developing, its lifespan increases and the 70% figure of L70 could reach the duration exceeding 60,000 hours. This leads the industry to adopt additional ratings, such as L80 (80% of initial lumens) and L90 (90% of initial lumens). In other words, a LED rated at 50,000 hours L90 would retain 90% of its light output at 50,000 hours, whereas a LED rated at 50,000 hours L70 would only retain 70% of its original output.
As LEDs continue to develop, their performance and quality keep improving. And your next LED flashlight can be more durable and reliable.
(The abrupt failure of incandescent sources in brown, and gradual dimming down of LED sources)
L70 is a pretty poor rating. You’re losing nearly 1/3 the light output. In contrast, T8 fluorescent has a 95% lumen maintenance over its lifespan whereas HPS is near 90%. Some municipal lighting manufacturers in Europe are using the more stringent L80 and L90 rating. Personally, if I were specifying the lighting, I would like an L90 rating.
What this chart also doesn’t tell us is the chromaticity shift over time. Some manufacturers will not cover or recognize color shifting in the warranty. So the lights start shifting green at 3000 hours, and you’re out of luck. I know the canopy lighting at the Circle K’s in Tucson, AZ have some pretty bad chromatic shifting towards the green. They were the early adopters, and they paid the price.
Curoius…
By that 5.7 year/24 hours a day metric, how could the LED manufacturers accurately rate their designs for L70/80/90?
By that metric, any lifespan rating for any LED newer than 5.7 years is just based on speculation… correct?
Do they actually test product for L70/80/90 under accelerated conditions? If so, what are the actual conditions used?
If I may have a guess, it’s probably extrapolation. A known decay function could be fitted to the data and calculated for a time point in the future. Or application of a high energy dose rate and compensation of the data for the acceleration effect of that higher-than-typical rate. Or both.
That could certainly be the case. You would need a spectrometer capable of measuring duv, and use that over time. Or you could compare the output of an old emitter vs a new one. It’s not obvious with LED when color shift happens over time, unless you see a new emitter next to the old or something else catastrophic has happened like massive LLD. Case in point, Berkley, CA , where steep light lumen depreciation combined with substantial chromaticity shift resulted in the premature replacement of LED street lights. The old lights were 4000K and looked pretty close to Oakland’s lights, but in the picture you can see they look a lot more yellow. It was obvious the lights in Berkley were changing color, and people noticed. However, things ended up on a positive note for Berkley as they got an upgrade to brand new 3000K lights.
Like hard drive specs. 100,000 hours MTBF. 14 year rating for a drive that’s been in production for 3 months?
Extrapolation, wishful thinking, or marketing?
All the Best,
Jeff
If have two of the original C01 in ww. The one in daily use is more rosy than any 219B. It is a reading light with probably 30 mins daily use. There was a slight difference from day 1, but it got extreme.
IMHO there is one other significant factor. Unlike the incandescent light source, LED’s require some form of electronic control in order to emit light. I have a number of LED’s that have effectively failed LONG before their expected life span - many within the expected life span of incandescent lights. On one hand I am nearly 100% certain that indeed the LEDs have not failed and are capable of emitting light perhaps to or exceeding their expected life. Yet, as used, the LED’s are not able to emit light because the electronic controls powering them have failed. Pretty hard for me to accept the idea of a useful life span when in reality the useful life span can not be achieved unless other required parts have that same useful lifespan.
Speaking of other factors that affect the lifespan of LEDs, has anyone thought of the thermal paste drying out over time? I worry that one day 4-6 years from now my favorite lights will need new TIM compound.
My monitor’s had a lot of powered on time (50-80k hrs), and it’s blue spiking so hard that even calibration can’t entirely fix it. It’s a Pro type monitor, so the LED were above average spectrum quality to begin with.
If users are accepting of a brand new “1000 lumen” or whatever light that will only sustain that for 1 minute due to thermal throttling, or even sustain only <50% of a higher normal mode for prolonged periods, is long term lumen maintenance really that much of a concern?
Let’s be honest, if there is a desire for honesty and truth in flashlight ratings, there is a lot more low hanging fruit ripe for picking before the lumen maintenance rating of an emitter is plucked from the tree.
I think the 50,000 hour lifespan is more applicable to LED home lights. These run for hours, not just minutes at a time. I dont think you can ever put 50,000 hours on a flashlight with normal use. Even by using a flashlight for one hour per day every day , would take 137 years for the LED to reach 70% brightness. Even the anodization on the flashlight would be flaking off by then.
It is possible, but the primary issue is that lights aren’t being rated honestly and transparently and advertised that way, even by legitimate manufacturers that don’t play the “our 18650 has 10,000 mAh” type of games.
Joe Six Pack consumer shopping for a light on Amazon isn’t going to be aware that the “1000 lumen” light isn’t going to produce that for the 60 minutes as stated in the Turbo column of the output table, never mind 6 minutes; more like 60 seconds. That’s pure fantasy, dishonest, and a deceptive way to market a product, but sadly, the norm.
Why that practice is allowed in the ANSI charts without even so much as an asterisk or some sort of disclaimer is puzzling.
Olight is on the receiving end of a lot of slings and arrows, but off the top of my head, they’re at least one company that makes an effort to be more honest in breaking down the output level and duration in their specs. Its top-level output specs clearly contain more than one set of figures, referencing both the short-burst and sustained performance at that level.
It matters because the industry is not producing LEDs with flashlights in mind but for general lighting where the LEDs are driven hard and hot and many hours per day.
