I have one test going on, the "10 yr shelf life test". Since I am 12, I think I might be able to do this until I am 22. Also, I want to test the life of a luxeon and actually having it die of age. so, how should I do this? someone before stated 100,000 hours is over 11 years. Any ideas?
How should I do 100,000 hr life of luxeon test?
- Thread starter Mags
- Start date
Help Support Candle Power Flashlight Forum
[ QUOTE ]
Mags said:
I have one test going on, the "10 yr shelf life test". Since I am 12, I think I might be able to do this until I am 22. Also, I want to test the life of a luxeon and actually having it die of age. so, how should I do this? someone before stated 100,000 hours is over 11 years. Any ideas?
[/ QUOTE ]
100,000 hours to 70% brightness I think is under IDEAL conditions. If a light is driven to high levels and/or has poor heatsinking, it won't go no where near 100,000 hours.
The new HDS EDC lights that use a Lux 3 are well made, and yet the owner Henry recently stated if memory of mine works life of 6,000 hours to 70% brightness when on high (60 lumens). So, 100,000 hours is most likely under ideal conditions with top notch heat sinking and driven at or under specs?
Mags said:
I have one test going on, the "10 yr shelf life test". Since I am 12, I think I might be able to do this until I am 22. Also, I want to test the life of a luxeon and actually having it die of age. so, how should I do this? someone before stated 100,000 hours is over 11 years. Any ideas?
[/ QUOTE ]
100,000 hours to 70% brightness I think is under IDEAL conditions. If a light is driven to high levels and/or has poor heatsinking, it won't go no where near 100,000 hours.
The new HDS EDC lights that use a Lux 3 are well made, and yet the owner Henry recently stated if memory of mine works life of 6,000 hours to 70% brightness when on high (60 lumens). So, 100,000 hours is most likely under ideal conditions with top notch heat sinking and driven at or under specs?
Ha ha ha ha ha! That's a good one Magzzie, and I applaud your diligence and energy.
Lessee now, ummm, 100,000 divided by 24 equals 4,125 days, divided by 7 equals about 590 weeks, divided by 52 equals about eleven-and-a-third YEARS...so, have a truckload of batteries nearby and as soon as it dims to some preset brightness (50% by light meter? You'd also have to buy a light meter, then) you change out the battery/ies.
Keep doing this through high school and the first few years of college, and yer done!
Report back to the Shady Valley Rest Home where I'll be stayin'...
Seriously, I really appreciate your even thinking about doing this. Yer a Flashoholic fer shore, right down to the very core... /ubbthreads/images/graemlins/grinser2.gif
Lessee now, ummm, 100,000 divided by 24 equals 4,125 days, divided by 7 equals about 590 weeks, divided by 52 equals about eleven-and-a-third YEARS...so, have a truckload of batteries nearby and as soon as it dims to some preset brightness (50% by light meter? You'd also have to buy a light meter, then) you change out the battery/ies.
Keep doing this through high school and the first few years of college, and yer done!
Report back to the Shady Valley Rest Home where I'll be stayin'...
Seriously, I really appreciate your even thinking about doing this. Yer a Flashoholic fer shore, right down to the very core... /ubbthreads/images/graemlins/grinser2.gif
[ QUOTE ]
Mags said:
I have one test going on, the "10 yr shelf life test". Since I am 12, I think I might be able to do this until I am 22. Also, I want to test the life of a luxeon and actually having it die of age. so, how should I do this? someone before stated 100,000 hours is over 11 years. Any ideas?
[/ QUOTE ]
100,000 hours to 70% brightness I think is under IDEAL conditions. If a light is driven to high levels and/or has poor heatsinking, it won't go no where near 100,000 hours.
The new HDS EDC lights that use a Lux 3 are well made, and yet the owner Henry recently stated if memory of mine works life of 6,000 hours to 70% brightness when on high (60 lumens). So, 100,000 hours is most likely under ideal conditions with top notch heat sinking and driven at or under specs?
Oh, and how do YOU test for led life? I hope you have lots of money for batteries. Testing out to 100,000 hours will consume lots of batteries, and if its 123 batteries you'll be down lots of money (even with deals on 123s of $1 per battery at Battery STation).
Mags said:
I have one test going on, the "10 yr shelf life test". Since I am 12, I think I might be able to do this until I am 22. Also, I want to test the life of a luxeon and actually having it die of age. so, how should I do this? someone before stated 100,000 hours is over 11 years. Any ideas?
[/ QUOTE ]
100,000 hours to 70% brightness I think is under IDEAL conditions. If a light is driven to high levels and/or has poor heatsinking, it won't go no where near 100,000 hours.
The new HDS EDC lights that use a Lux 3 are well made, and yet the owner Henry recently stated if memory of mine works life of 6,000 hours to 70% brightness when on high (60 lumens). So, 100,000 hours is most likely under ideal conditions with top notch heat sinking and driven at or under specs?
Oh, and how do YOU test for led life? I hope you have lots of money for batteries. Testing out to 100,000 hours will consume lots of batteries, and if its 123 batteries you'll be down lots of money (even with deals on 123s of $1 per battery at Battery STation).
KahveOrta
Newly Enlightened
Re: How should I do 100,000 hr life of luxeon test
Ha...that's a good one, Mags. I really admire your enthusiasm
Instead of running off batteries, mod a luxeon to run off the mains and use it as a nightlight (NOT that I'm implying that you need one, but I do have a son about your age that can't sleep without a nightlight!)
Good luck
Ha...that's a good one, Mags. I really admire your enthusiasm
Instead of running off batteries, mod a luxeon to run off the mains and use it as a nightlight (NOT that I'm implying that you need one, but I do have a son about your age that can't sleep without a nightlight!)
Good luck
HDS_Systems
Enlightened
Mags and 357,
There are two types of failures that generally apply to LEDs when one asks about life. The first type of failure is a catastrophic failure, such as an open or short. These are rare - especially if you do not abuse the LED. Typical mean time to failure (MTTF) is better than 150,000 hours under proper conditions - the typical "life" you see quoted by many manufacturers.
The second type of failure is loss of lumen output. This is an aging function which takes place in all LEDs. Heat and current density tend to be the driving forces while material degradation tends to be the actual cause. Most manufacturers rate the life of an LED based on the loss of 30% of the initial lumen output under some specific set of conditions. Of course, the set of conditions selected by marketing to state a claim and the set of conditions the LED will actually see in use may be quite different. 5,000 to 20,000 hours is common and for most practical purposes is greater than the useful life of the product. In other words, the device will be taken out of service for other practical reasons long before you exceed the useful life of the LED.
For instance, on the 60 lumen setting with an EDC-series flashlight you should easily exceed 5,000 hours life on the LED, which is roughly 15,000 battery changes. If you drop the output power down to 10 lumens, you can probably get something closer to 50,000 hours of life from the same LED, which is roughly 5000 battery changes. At the 2 lumen setting, you can probably exceed 150,000 hours from the same LED, which is still over 1000 battery changes. Suffice it to say that in our products, the LED is good for the life of the product. These numbers tend to be conservative.
Remember, life is primarily a function of current and heat. Vendors who over drive their LEDs will see significantly shorter life times - including reduced MTTF - sometimes less than 1000 hours. It all depends on the actual operating conditions. Don't expect an LED with one plus amps running through it to last a long time without some serious cooling attached.
Henry.
There are two types of failures that generally apply to LEDs when one asks about life. The first type of failure is a catastrophic failure, such as an open or short. These are rare - especially if you do not abuse the LED. Typical mean time to failure (MTTF) is better than 150,000 hours under proper conditions - the typical "life" you see quoted by many manufacturers.
The second type of failure is loss of lumen output. This is an aging function which takes place in all LEDs. Heat and current density tend to be the driving forces while material degradation tends to be the actual cause. Most manufacturers rate the life of an LED based on the loss of 30% of the initial lumen output under some specific set of conditions. Of course, the set of conditions selected by marketing to state a claim and the set of conditions the LED will actually see in use may be quite different. 5,000 to 20,000 hours is common and for most practical purposes is greater than the useful life of the product. In other words, the device will be taken out of service for other practical reasons long before you exceed the useful life of the LED.
For instance, on the 60 lumen setting with an EDC-series flashlight you should easily exceed 5,000 hours life on the LED, which is roughly 15,000 battery changes. If you drop the output power down to 10 lumens, you can probably get something closer to 50,000 hours of life from the same LED, which is roughly 5000 battery changes. At the 2 lumen setting, you can probably exceed 150,000 hours from the same LED, which is still over 1000 battery changes. Suffice it to say that in our products, the LED is good for the life of the product. These numbers tend to be conservative.
Remember, life is primarily a function of current and heat. Vendors who over drive their LEDs will see significantly shorter life times - including reduced MTTF - sometimes less than 1000 hours. It all depends on the actual operating conditions. Don't expect an LED with one plus amps running through it to last a long time without some serious cooling attached.
Henry.
[ QUOTE ]
HDS_Systems said:
The second type of failure is loss of lumen output. This is an aging function which takes place in all LEDs. Heat and current density tend to be the driving forces while material degradation tends to be the actual cause. Most manufacturers rate the life of an LED based on the loss of 30% of the initial lumen output under some specific set of conditions. Of course, the set of conditions selected by marketing to state a claim and the set of conditions the LED will actually see in use may be quite different. 5,000 to 20,000 hours is common and for most practical purposes is greater than the useful life of the product. In other words, the device will be taken out of service for other practical reasons long before you exceed the useful life of the LED.
[/ QUOTE ]
Excellent explanation. Even at 5,000 hours, that is so much time that probably 99.99% of the users never even come close to reaching that point. That is plenty of led life, and also given that 10 lumens is plenty of light for most every-day tasks, I don't think anyone would ever wear out an HDS solely due to led light output loss.
HDS_Systems said:
The second type of failure is loss of lumen output. This is an aging function which takes place in all LEDs. Heat and current density tend to be the driving forces while material degradation tends to be the actual cause. Most manufacturers rate the life of an LED based on the loss of 30% of the initial lumen output under some specific set of conditions. Of course, the set of conditions selected by marketing to state a claim and the set of conditions the LED will actually see in use may be quite different. 5,000 to 20,000 hours is common and for most practical purposes is greater than the useful life of the product. In other words, the device will be taken out of service for other practical reasons long before you exceed the useful life of the LED.
[/ QUOTE ]
Excellent explanation. Even at 5,000 hours, that is so much time that probably 99.99% of the users never even come close to reaching that point. That is plenty of led life, and also given that 10 lumens is plenty of light for most every-day tasks, I don't think anyone would ever wear out an HDS solely due to led light output loss.
VidPro
Flashlight Enthusiast
which color?
from what i read, the current should be controlled on them to exact specs, especially when you run the full voltage specs it is often over on the current.
you would probably want to run more than 1, so you can tell if it is (like said above) catrostrofic.
They dont say 100,000 anymore, i read one of thier rescent PDF and it says 50,000, so you may get this done early enough.
but i say 6000 for the whites, to do the 80% thing WHY? because that is about how long phosphors are lasting in the rest of the stuff. mabey 20,000 before you would want to replace it.
but i have already run a set of 1w whites for way more than 6000, and its over 80%, but only 1/2 the time at spec.
and the indicator lights on components i have seen run for 20 years, so the led chemistry is capable of it.
what will happen when the phosphors are mostly fried? will the blue shine on?
from what i read, the current should be controlled on them to exact specs, especially when you run the full voltage specs it is often over on the current.
you would probably want to run more than 1, so you can tell if it is (like said above) catrostrofic.
They dont say 100,000 anymore, i read one of thier rescent PDF and it says 50,000, so you may get this done early enough.
but i say 6000 for the whites, to do the 80% thing WHY? because that is about how long phosphors are lasting in the rest of the stuff. mabey 20,000 before you would want to replace it.
but i have already run a set of 1w whites for way more than 6000, and its over 80%, but only 1/2 the time at spec.
and the indicator lights on components i have seen run for 20 years, so the led chemistry is capable of it.
what will happen when the phosphors are mostly fried? will the blue shine on?
Mike Painter
Flashlight Enthusiast
- Joined
- Sep 16, 2002
- Messages
- 1,863
Wait until you can find a 110 volt LED.
Plug it in in a garage or hallway.
Wait.
Plug it in in a garage or hallway.
Wait.
jtr1962
Flashaholic
I've had a Luxeon running at 350 mA on a small heatsink since last January using a wall wart for power. Vf dropped quickly from the initial 3.29V and has now stabilized at 2.91V. Brightness at one meter is 10.4 lux last time I checked. I have the light on my workroom ceiling where it serves as a nightlight of sorts. I'll let everyone know when it reaches 70% of initial brightness. This should be in about 4.5 to 5 years if Lumiled's data is accurate.
I don't expect that the LED will ever quit completely, just get dimmer and dimmer. I have some electronic devices with indicator LEDs which are still working after 20+ years. My best guess is that a properly driven and protected LED should last 200,000 hours or more, so I'll probably be a senior citizen before my Luxeon quits. I'm still using an alarm clock with a vacuum fluorescent display which has been in continuous use since 1978. That's over 235,000 hours. Properly designed electronic devices can last for generations, possibly even for centuries. Nobody really knows yet for sure.
The sad part of all this is that most electronic devices become obsolete long before they're non-functional. LEDs are no different but once efficiency reaches close to its theoretical limits they'll likely be used until they fail, rather than being replaced with better models long before.
I don't expect that the LED will ever quit completely, just get dimmer and dimmer. I have some electronic devices with indicator LEDs which are still working after 20+ years. My best guess is that a properly driven and protected LED should last 200,000 hours or more, so I'll probably be a senior citizen before my Luxeon quits. I'm still using an alarm clock with a vacuum fluorescent display which has been in continuous use since 1978. That's over 235,000 hours. Properly designed electronic devices can last for generations, possibly even for centuries. Nobody really knows yet for sure.
The sad part of all this is that most electronic devices become obsolete long before they're non-functional. LEDs are no different but once efficiency reaches close to its theoretical limits they'll likely be used until they fail, rather than being replaced with better models long before.
PeLu
Flashlight Enthusiast
Of course such a test could be done, but it has to be as many samples as possible (lets say 100) to get statistically sigificant data. Temp and other values have to be within a specific range and then, 11 years later it will turn out that nobody is actually interested about the lifetime and behaviour of a LED made in 2005 .-)
Yes, maybe a few CPFers (or how we might be called then) would give it a few seconds of their time and start talking about the good old times.
Yes, maybe a few CPFers (or how we might be called then) would give it a few seconds of their time and start talking about the good old times.
PeterW
Enlightened
Re: How should I do 100,000 hr life of luxeon test
Do you think Limileds have actually waited11yrs to come to market... just think of the stuff they have just started testing!!
Seriously, to get their data, they have run the LEDs at high temps and looked at the deathrate. By doing this at a couple of high temps it is possible to determine an aging parameter that allows you to predict the deathrate for other temperatures. The lifetime for higher temperatures rolls off very rapidly..... I think I remember a colleague saying half the life for +10degC, so Lumileds just pop the LED's in a furnace and wait for the pop, muliply the time by a few thousands etc and there you go. There is a lot of work being done in getting the predictions more accurately.
Cheers
PeterW
Do you think Limileds have actually waited11yrs to come to market... just think of the stuff they have just started testing!!
Seriously, to get their data, they have run the LEDs at high temps and looked at the deathrate. By doing this at a couple of high temps it is possible to determine an aging parameter that allows you to predict the deathrate for other temperatures. The lifetime for higher temperatures rolls off very rapidly..... I think I remember a colleague saying half the life for +10degC, so Lumileds just pop the LED's in a furnace and wait for the pop, muliply the time by a few thousands etc and there you go. There is a lot of work being done in getting the predictions more accurately.
Cheers
PeterW
[ QUOTE ]
Mags said:
I have one test going on, the "10 yr shelf life test". Since I am 12, I think I might be able to do this until I am 22. Also, I want to test the life of a luxeon and actually having it die of age. so, how should I do this? someone before stated 100,000 hours is over 11 years. Any ideas?
[/ QUOTE ]
If you could get your dog to run this test, it should only take about 1.5 years.
Mags said:
I have one test going on, the "10 yr shelf life test". Since I am 12, I think I might be able to do this until I am 22. Also, I want to test the life of a luxeon and actually having it die of age. so, how should I do this? someone before stated 100,000 hours is over 11 years. Any ideas?
[/ QUOTE ]
If you could get your dog to run this test, it should only take about 1.5 years.
Finbar
Newly Enlightened
- Joined
- Dec 15, 2004
- Messages
- 191
Re: How should I do 100,000 hr life of luxeon test
You could overdrive the LED and make an inference about the predicted life span.
Run several LED's at different levels of overdrive and make a predictive graph of the data.
You would not need a lot of batteries. I read an article about a LEd being driven from a 12V car battery. You could have a setup to run off of DC current from the wall socket. Just need to calculate the proper resistors to run it. Cool some with a computer fan some without. Overdrive some and others do not. Get a big chunck of metal for a heat sink and run the test.
Nice to make a LED lamp.
Hook one up in the freezer.
It would be nice to have an independent test done by someone other than the seller of LED's.
Shine on you crazy diamond,
Fin
You could overdrive the LED and make an inference about the predicted life span.
Run several LED's at different levels of overdrive and make a predictive graph of the data.
You would not need a lot of batteries. I read an article about a LEd being driven from a 12V car battery. You could have a setup to run off of DC current from the wall socket. Just need to calculate the proper resistors to run it. Cool some with a computer fan some without. Overdrive some and others do not. Get a big chunck of metal for a heat sink and run the test.
Nice to make a LED lamp.
Hook one up in the freezer.
It would be nice to have an independent test done by someone other than the seller of LED's.
Shine on you crazy diamond,
Fin
