How long will an overdriven luxeon last?
- Thread starter zmoz
- Start date
Help Support Candle Power Flashlight Forum
Rothrandir
Flashaholic
depends on how well it's heatsunk /ubbthreads/images/graemlins/grin.gif
if no outside heatsinking is used...it won't last more than a couple seconds.
if amazing heatsinking is used, it could last longer than 100000hours /ubbthreads/images/graemlins/grin.gif
if no outside heatsinking is used...it won't last more than a couple seconds.
if amazing heatsinking is used, it could last longer than 100000hours /ubbthreads/images/graemlins/grin.gif
shipinretech
Enlightened
I doubt that it will last longer than rated on a consistent basis if overdriven regardless of the heatsinking. The thermal event occuring on the die will be substantially greater than the technology is capable of sustaining. Every microscopic flaw of the die will be a source of spreading failure. The effect of physical shocks will be much greater and be much more likely to cause fatal delaminations. The power supply should be incredibly stable because overvoltage variances are another source of catastrophic failure.
I doubt that there have been many comprehensive tests of overdriving failure rates of Luxeon LEDs outside of Lumileds because nobody has deep enough pockets. Most of the informal "tests" I have seen have a disturbing resemblence to the thermal testing of insects with magnifying glasses by juvenile males. The primary goal appears to be the gratification of watching the LED turn to smoke.
Practially speaking, heatsinking is always good, but you should never depend on a light that is overdriven. The likelihood of failure is great and the moment of failure is impossible to determine. Betting your life on a MTBF failure curve that is unknown is a bad plan.
I doubt that there have been many comprehensive tests of overdriving failure rates of Luxeon LEDs outside of Lumileds because nobody has deep enough pockets. Most of the informal "tests" I have seen have a disturbing resemblence to the thermal testing of insects with magnifying glasses by juvenile males. The primary goal appears to be the gratification of watching the LED turn to smoke.
Practially speaking, heatsinking is always good, but you should never depend on a light that is overdriven. The likelihood of failure is great and the moment of failure is impossible to determine. Betting your life on a MTBF failure curve that is unknown is a bad plan.
Rothrandir
Flashaholic
i think the life of the luxeon is very much related to thermal management.
active heatsinking (freezer for example) would allow the die to be substantially cooler than normal operating temperatures require, and as such, prolong life.
most of my mods are overdriving, and while that is certainly detremintal to the luxeons health compared to driving at specs or underdriving, i use adaquate (overkill) heatsinking, and as a result are very confident.
furthermore, although i am never without at least 3 lights, i would be fully confident to rely on any of my mods as much as i rely on any other light. i also believe that if one of them does "fail" it will be a result of some other force then overheating the luxeon, such as batteries, switch, elephant herd, etc...
active heatsinking (freezer for example) would allow the die to be substantially cooler than normal operating temperatures require, and as such, prolong life.
most of my mods are overdriving, and while that is certainly detremintal to the luxeons health compared to driving at specs or underdriving, i use adaquate (overkill) heatsinking, and as a result are very confident.
furthermore, although i am never without at least 3 lights, i would be fully confident to rely on any of my mods as much as i rely on any other light. i also believe that if one of them does "fail" it will be a result of some other force then overheating the luxeon, such as batteries, switch, elephant herd, etc...
This definately won't be something to depend on. I'm just wondering about how long I can expect before it burns out. /ubbthreads/images/graemlins/smile.gif Is it something I should worry about, or is it so long that it really doesn't matter?
Rothrandir
Flashaholic
i alread told you it was something you should worry about /ubbthreads/images/graemlins/twak.gif
very nice though /ubbthreads/images/graemlins/grin.gif
very nice though /ubbthreads/images/graemlins/grin.gif
Lets just say, within reason is what you should expect to drive it if you want the LS to last within reasonably stated specs... I doubt no one's ever going to know the death of an LS at a given overdrive, but I'd figure if it survives being driven hard and is still productive for maybe 500 hours, it's all good...
A turbocharged engine will not last as long as an engine carefully tuned to lope along sanely and slowly, performing well within its limits.
And those are the types of cars I passed every day in my tricked out '87 Buick Grand National -- five years and 106,000 miles later I gave it, still running perfectly, to my nephew.
Depends how you want to live your life I guess, and the philosophy that goes along with that kind of thinking.
In financial affairs I have avoided mutual funds and the stock market, preferring to go with Universal Life and triple-A rated bonds, but we're talking flashlight hobbying here, and I try to keep things in perspective.
And those are the types of cars I passed every day in my tricked out '87 Buick Grand National -- five years and 106,000 miles later I gave it, still running perfectly, to my nephew.
Depends how you want to live your life I guess, and the philosophy that goes along with that kind of thinking.
In financial affairs I have avoided mutual funds and the stock market, preferring to go with Universal Life and triple-A rated bonds, but we're talking flashlight hobbying here, and I try to keep things in perspective.
The more you push the LED above the rated input current, the brighter the light output. /ubbthreads/images/graemlins/ooo.gif
The brighter the light output, the hotter the LED gets. /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/ooo.gif
The hotter the LED gets, the faster it deteriorates. /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/ooo.gif
The faster it deteriorates, the shorter it lives. /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/ooo.gif
So, in short, enjoy it while it lasts if you plan to overdrive it. /ubbthreads/images/graemlins/naughty.gif
But.... what a way to go /ubbthreads/images/graemlins/grin.gif
The brighter the light output, the hotter the LED gets. /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/ooo.gif
The hotter the LED gets, the faster it deteriorates. /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/ooo.gif
The faster it deteriorates, the shorter it lives. /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/ooo.gif /ubbthreads/images/graemlins/ooo.gif
So, in short, enjoy it while it lasts if you plan to overdrive it. /ubbthreads/images/graemlins/naughty.gif
But.... what a way to go /ubbthreads/images/graemlins/grin.gif
shipinretech
Enlightened
Rothrandir,
You are a flashaholic among flashaholics and your memory will last forever; unlike your overdriven Luxeons. Heck, in three or four years, about the time most overdriven, underused are letting the magic smoke get away, we should be getting more light out of underdriven LEDs than we can get out of overdriven LEDs today. I do agree with overheatsinking high-power LEDs regardless of the power used, but experience with extreme environments makes me like over-engineered equipment.
As long as you have backup to go with your overdriven light and you know that you have an overdriven light, it's all good. A lot of the 2 for $20 import lights are very overdriven and the consumer does not know it. That bugs me.
Mr. Bulk,
I respect the endurance of your creation and the skill of your nephew in keeping it between the lines. You live in Hawaii, and keeping it between the lines with that much power on twisty roads is a bigger accomplishment than doing so in Texas or Oklahoma. Also unlike Texas and New Mexico, your state is not particularly known for dust storms or dirt roads, two leading causes of turbo failure. In the same vein, while overdriven lights may work fine in some environments, other more hostile environments may deliver different results.
You are a flashaholic among flashaholics and your memory will last forever; unlike your overdriven Luxeons. Heck, in three or four years, about the time most overdriven, underused are letting the magic smoke get away, we should be getting more light out of underdriven LEDs than we can get out of overdriven LEDs today. I do agree with overheatsinking high-power LEDs regardless of the power used, but experience with extreme environments makes me like over-engineered equipment.
As long as you have backup to go with your overdriven light and you know that you have an overdriven light, it's all good. A lot of the 2 for $20 import lights are very overdriven and the consumer does not know it. That bugs me.
Mr. Bulk,
I respect the endurance of your creation and the skill of your nephew in keeping it between the lines. You live in Hawaii, and keeping it between the lines with that much power on twisty roads is a bigger accomplishment than doing so in Texas or Oklahoma. Also unlike Texas and New Mexico, your state is not particularly known for dust storms or dirt roads, two leading causes of turbo failure. In the same vein, while overdriven lights may work fine in some environments, other more hostile environments may deliver different results.
Rothrandir
Flashaholic
patrick.
what you say is true, and i totally agree.
what i was saying in my first post in this thread, is that if you overdrive a luxeon, and still keep it as cool or cooler than a normally driven luxeon, you can get the same or better life.
usually, this would require some sort of active cooling...
it is also relative....not matter how cool you keep the luxoen, it's not going to like being driven with 50volts. i am speaking in perspective to how a luxeon is normally overdriven.
even if moderately overdriving, with heatsinking intended for spec-driving, i can assume that a light would outlast me /ubbthreads/images/graemlins/grin.gif
most led lights out on the market that are being grossly overdriven are not led-based, and are much more picky about thermal issues. i have no problem with overdriving a luxeon, but am very conservative when it comes to overdriving 3 or 5mm leds.
what you say is true, and i totally agree.
what i was saying in my first post in this thread, is that if you overdrive a luxeon, and still keep it as cool or cooler than a normally driven luxeon, you can get the same or better life.
usually, this would require some sort of active cooling...
it is also relative....not matter how cool you keep the luxoen, it's not going to like being driven with 50volts. i am speaking in perspective to how a luxeon is normally overdriven.
even if moderately overdriving, with heatsinking intended for spec-driving, i can assume that a light would outlast me /ubbthreads/images/graemlins/grin.gif
most led lights out on the market that are being grossly overdriven are not led-based, and are much more picky about thermal issues. i have no problem with overdriving a luxeon, but am very conservative when it comes to overdriving 3 or 5mm leds.
On a worst case scenario of an overdriven LED lasting only approximately 5000 hours against 50,000 to 100,000 of a undriven LED the average user would, using a light for 6 hours a night continuously for 7 days a week would get about two and a half years of use out of the LED.
Now all being equal most of us dont use a light for anything like 6 hours a night or for 365 days a year. The probability being that we might therefore get up to anything like four to five years of use is IMO exceptionally good. In four or five years time I am in no doubt that LED technology will have progressed far beyond what we see (pun) in use now, and we, as flashaholics will be enjoying something new to play with.
I cant get steamed up over whether an overdriven LED will die young as I have enough normally driven LEDs giving me stirling service and I know that I will have replaced them anyway long before they are even near their half year life.
I depend on my lights at work and I nearly always have at least two on me at nights (my own purchases) and there are two others in my response car (service issue lights).
Overdriving LEDs is taking a leap of faith, for most folk it is an exceptable risk as the potential for a sudden LED death is rare. You pays your money and takes your choice? Your call?
Doug
Now all being equal most of us dont use a light for anything like 6 hours a night or for 365 days a year. The probability being that we might therefore get up to anything like four to five years of use is IMO exceptionally good. In four or five years time I am in no doubt that LED technology will have progressed far beyond what we see (pun) in use now, and we, as flashaholics will be enjoying something new to play with.
I cant get steamed up over whether an overdriven LED will die young as I have enough normally driven LEDs giving me stirling service and I know that I will have replaced them anyway long before they are even near their half year life.
I depend on my lights at work and I nearly always have at least two on me at nights (my own purchases) and there are two others in my response car (service issue lights).
Overdriving LEDs is taking a leap of faith, for most folk it is an exceptable risk as the potential for a sudden LED death is rare. You pays your money and takes your choice? Your call?
Doug
shipinretech
Enlightened
dougmccoy,
While your premise is theoretically sound, there are two exceptions to it.
First, the amount of overdriving typical in the market is a factor of three with 5mm. For on topic matters, you are correct because most Luxeons are being overdriven a fraction of their rated power, but in general, 5mm LEDs are typically overdriven by a multiple.
Second, because of the above, the lifespan changes from rated life to mean time between failure (MTBF). The difference is that MTBF means that most products will last the given span, some will last longer, and some will fail the next time they are used. The MTBF of a multiple overdriven LED is closer to 1,000 hours, 20% of the life you claim. Worse than that is the problem that the MTBF is predicated on optimum conditions, something rarely found in survival situations.
I get steamed because I see overdriven LEDs being marketed as having 100K hours of life with the implied promise that they will not fail when they are designed specifically to do so. Since these products have huge numbers, that means they are failing on a regular basis. That not only makes the LED lighting industry look bad, it puts us at risk of serious liability from rapacious tort lawyers. Although it is unlikely, this industry could be strangled in the crib by frivolous lawsuits. If somebody dies and they happen to have an overdriven Luxeon as part of their death, a tort lawyer could sue Lumileds for millions and trash the industry. Think it couldn't happen? As a developer of commercial products, I'm not ready to take that leap of faith.
While your premise is theoretically sound, there are two exceptions to it.
First, the amount of overdriving typical in the market is a factor of three with 5mm. For on topic matters, you are correct because most Luxeons are being overdriven a fraction of their rated power, but in general, 5mm LEDs are typically overdriven by a multiple.
Second, because of the above, the lifespan changes from rated life to mean time between failure (MTBF). The difference is that MTBF means that most products will last the given span, some will last longer, and some will fail the next time they are used. The MTBF of a multiple overdriven LED is closer to 1,000 hours, 20% of the life you claim. Worse than that is the problem that the MTBF is predicated on optimum conditions, something rarely found in survival situations.
I get steamed because I see overdriven LEDs being marketed as having 100K hours of life with the implied promise that they will not fail when they are designed specifically to do so. Since these products have huge numbers, that means they are failing on a regular basis. That not only makes the LED lighting industry look bad, it puts us at risk of serious liability from rapacious tort lawyers. Although it is unlikely, this industry could be strangled in the crib by frivolous lawsuits. If somebody dies and they happen to have an overdriven Luxeon as part of their death, a tort lawyer could sue Lumileds for millions and trash the industry. Think it couldn't happen? As a developer of commercial products, I'm not ready to take that leap of faith.
Hi guys,
I recently received a new estimate of life from Nichia regarding their NSPW white LED's and the half life of these guys has gone down from 80,000 to 8,000 to 10,000 and ins some cases of higher temp and humidity, the half life is down below 5000 hours. These expectations are based on the current at a max of 20 mA.
Anyone claiming 100,000 hours is dreaming, it would now appear and in the case of overdriven LED's like Patrick has mentioned, it may well be the case that flashlight owners *will* see their lights go dim during what many would consider normal use and life of the light.
If a light will degrade by a half life during its use, one could argue that reducing the drive current will give an initially dimmer light that won't be seen to go dim during it's expected usage cycle. I suspect that we have some lights in our population that have half lifes in the low hundreds of hours and even less. As time goes buy, these lights might start to show up their limited span.
- Don
I recently received a new estimate of life from Nichia regarding their NSPW white LED's and the half life of these guys has gone down from 80,000 to 8,000 to 10,000 and ins some cases of higher temp and humidity, the half life is down below 5000 hours. These expectations are based on the current at a max of 20 mA.
Anyone claiming 100,000 hours is dreaming, it would now appear and in the case of overdriven LED's like Patrick has mentioned, it may well be the case that flashlight owners *will* see their lights go dim during what many would consider normal use and life of the light.
If a light will degrade by a half life during its use, one could argue that reducing the drive current will give an initially dimmer light that won't be seen to go dim during it's expected usage cycle. I suspect that we have some lights in our population that have half lifes in the low hundreds of hours and even less. As time goes buy, these lights might start to show up their limited span.
- Don
