Do LEDs really last 10,000+ hours?

[Elysia]

I have a question about the reported life of LEDs. Has anyone actually experienced a LED lasting for tens of thousands of hours? When the LED moonlights went out on my combo compact fluor. hood, I though that the heat of the other lights was to blame. I purchased one of the Marineland LED lights and used it for about 4 months, for a maximum of 2,000 hours. Only to notice the other day that the "blue" LEDs are no longer lit. While the white LEDs still work, I feel that they are very yellow, so the entire fixture has little use anymore.
Do LEDs typically only last a fraction of their reported lifetime? I seem to have the same luck with CFL bulbs in regular fixtures -- I swear I have changed more CFL bulbs than incandescent, despite the fact that they are suppose to provide light for many more hours than an incandescent.

 

[bamf25]

With the CFL's it is becoming widely reported that they are way under performing with regards to lifespan.

 

[Dadekster]

Has anyone owned one long enough to know if they last as long as they say they are rated for? Isn't five years of moderate use considered normal? I am curious about this as well tbh. I have also read about issues with the CFL's and hope that this is not standard across all LED lighting. I doubt it is, but you never know these days.

 

[Anemonebuff]

They are rated to last from 30,000 to 100,000 hours of use under optimal circumstances. Cooled properly and assuming that they were quality LEDs and mounted and powered properly, they will all fall into those ratings. These are electrical components and they are subject to failure due to normal numbers of defective parts/craftmanship. You also must remember that they are degrading during this time, just much slower than other forms of lighting. There are quite a few people in the 1,2,3+ years of LED use so far and I would guess there are some that have used them longer. However, they are not the same LEDs produced in the last year or so. CFLs are way over rated and the global warming crowd refuses to drop them and adopt LEDs and I do not know why, particularly because LEDs do not have mercury in them and last many years longer. I would say that todays LEDs, used properly will get you at least 5 years of good use.

 

[nanojg]

I think the OP was asking for actual experience, what you described is basically predictions (kinda like the CFL predictions). I hope you are right though, as I am sold on leds myself, even if they dont last the full 5 years or longer.

 

[Shard]

I think the OP was asking for actual experience, what you described is basically predictions (kinda like the CFL predictions).

What he is describing are not just wild 'predictions'. The expected life data is based on actual testing to generate a degradation curve. As long as the LEDs are mounted properly so they do not overheat and are applied the proper power profile, they will last as long as the datasheet says they will. It is really based in science, not opinion.

As for the OPs experience with his fixture, the white LEDs look yellow, because most fixtures tend to use ~6500k white LEDs. The light at that kelvin is yellowish, which isn't necessarily bad, because it is more conducive to coral growth than higher kelvin light. The blues offset the color to give it a 10-20k look.

 

[nanojg]

What he is describing are not just wild 'predictions'. The expected life data is based on actual testing to generate a degradation curve. As long as the LEDs are mounted properly so they do not overheat and are applied the proper power profile, they will last as long as the datasheet says they will. It is really based in science, not opinion.

As for the OPs experience with his fixture, the white LEDs look yellow, because most fixtures tend to use ~6500k white LEDs. The light at that kelvin is yellowish, which isn't necessarily bad, because it is more conducive to coral growth than higher kelvin light. The blues offset the color to give it a 10-20k look.

So are you saying that the CFL life spans were "wild predictions" and NOT based on actual testing? Also I disagree that in a reef environment that they will correspond with the datasheet, the datasheet specs were tested under optimal conditions, without daily on/offs and dimming.

 

[widmer]

I purchased one of the Marineland LED lights and used it for about 4 months, for a maximum of 2,000 hours. Only to notice the other day that the "blue" LEDs are no longer lit.

There are a couple things going on here.

1. It is quite possible that only one of the LEDs went out, and that either opened the circuit, or caused the rest of them to be overdriven and fried.

2. I question a lot of the manufactured LED fixtures that don't necessarily specify the characteristics of the LEDs that are built into them, or don't even name the manufacturer of the LEDs that were used. Like any other product, you can buy a quality-manufactured item that was built to last, or you can buy one that's "flimsy" and has a short lifespan.

Especially now with marineland's stealth heater catastrophe, I'm not saying that marineland must make poor quality products, but rather, this manufacturer is clearly capable of making something that will prematurely fail.

But truly not all LEDs are created equal. Just like incandescent bulbs, some were manufactured in such a way that they will last two or three times longer than others...

 

[ufoman420]

I built my DIY led light for my 170gal reef tank about a year and half ago and i love it. I have not changed anything i have about 1" 1/2 growth on my monipora and zoes spreading like crazy, my candy cane split from 4 heads to about 16 and my pipe organ doubled its size. I have 144 3 watt leds and my hard corals are about 6" below the water surface and no problems at all. My plate coral is at the bottom of the tank 24" and it has a great pink color and doing great.

 

[Pink_n_Purple]

All light sources ie compac flores, mh, hps, leds, no-ho-vho flores are all tested and rated in optimal conditions. Humidity, temp, balast, and even source of power 120'240 volts, electrical and magnetic interferance can all come into play when discussing effiency and life of your light source. Imo leds are still a new tech and havnt been around long enough in these situations (horticulture, aquariums) to even know how long they will last or perform. Not the to mention the mixing of parts, and poor building practice of these unheard of manafactures, buying a power supply made by a different company than the light source itself! like buying a hortilux 60hz bulb and running in on a lumitek ballast @ 22,000hz, expect it not to perform optimaly. All in all LED's r unpredictable, imo

 

[BFG]

As far as I can remember back in the day when led was breaking into the hobby, manufacturer were claiming that their led was able to operate for a few thousand hours at a low temperature of 20 degree celsius ( I might be wrong ). But only 1 led lightset manufacturer was disagreeing that it was impossible to operate a bank of led light at that low temperature without active cooling and that even with active cooling, it does not ensure all the led bulb are evenly cooled to lay their led bulb lifespan of a few thousand hours. My leds are coming to a year of usage and it seems to look ever so slightly dimmed but to me they still looks usable. I do not have any equipment to support my claim but just take it with a pinch of your favourite brand salts.

 

[ufoman420]

Mine run sort of hot when i touch the heat sink it is a little to warm to hold your hand too. I run my leds 13 hrs a day 8am to 9pm. It has been a year and a half and i have not noticed any difference in the output of light IE dimming or tank looking dull. I didn't use Cree's too expensive i used some china produced ones for 1.95 a piece and i would totally recommend them for use again. I did put 12 Cree's on there to see if I could visually see a difference and to my eye I do not see a difference. And price cost difference i could put 40 more led on my light for just the same price spent on Crees. To your point there have been tests done to conclude that the output is the same after 5,000 hours of use. But it is electronics and you will have the occasional bad one that goes out.

 

[snorvich]

Mean Time to Failure is normally what is quoted and this, of course implies a standard deviation. What this means is that some percentage of the time you will find failures (below the mean) or successes (above the mean) at the three standard deviation level from the mean which is 99%. If they provided the standard deviation you could estimate your own curve of failure and success probability.

 

[Pink_n_Purple]

i believe leds are rated for 50,000 hours of use, thats 10 years @ 12 on 12 off, ask anyone if there leds have lasted that long, either one burns out and comprimises the rest in the same seriese or the controller goes, for the price of them.... í'll stick to the old school mh,hps and ho flores

 

[Shard]

So are you saying that the CFL life spans were "wild predictions" and NOT based on actual testing?

No, that is not what I am saying. All I am saying is that the curves generated by the statistics are a fairly good representation of the life span of the LEDs because the expected matches the observed so far, so the probability is it will continue to do so in the future. Operating conditions are an implied factor in this of course.

Also I disagree that in a reef environment that they will correspond with the datasheet, the datasheet specs were tested under optimal conditions, without daily on/offs and dimming.

Depends on the situation.... Some LED setups like the Aqua Illumination modules have logic control on their internal temperature. Others may use thermal cutoff switches that keep the temperature from going outside of their operating range. Many could mimic datasheet conditions fairly well.

Most CFL fixtures do not have the level of internal thermal regulation/control that some LED setups do, so they may spend a lot of time outside of their datasheet operating/testing ranges. Of course, some cheap LED fixtures, or simple DIY setups would do the same thing. I guess I am just saying that it is possible to have various LED setups, as well as CFL setups that last for widely differing amounts of time.

If you re-read my post, you will see that I did say that temperature control and proper power profile are a key part of having them last as long as the datasheet says. I guess I could be more explicit and say the closer you mimic optimum conditions, the better your results will be.

As for the on/off situation, some do account for daily on/offs as many of the lifespan quotes out there are based on X hours per day and so on. As for maintaining optimal conditions, logic control or thermal regulation devices can keep it pretty close as mentioned above.

As for the dimming, LEDs are dimmed by changing the current that goes through the circuit. So if the projection curve was generated at 700ma, and you run yours at 350ma, you may get more time out of them, as they will run cooler, have less stress on them, etc.

Anyway, it is nice to see you like LEDs as well. I think they will become more common place in reefing in the future. If I could afford to go all LED now I would

Cheers,
Landon

 

[Shard]

Hi Pink_n_Purple, welcome to RC! I see you just joined.

Imo leds are still a new tech and havnt been around long enough in these situations (horticulture, aquariums) to even know how long they will last or perform.

Well if it helps, mine have lasted for a while, and there are others here with fixtures that have been going and growing for years. LED performance is documented over long courses of time here in some threads. Here is one that shows some actual SPS growth over time.

http://www.reefcentral.com/forums/showthread.php?t=1968168

Not the to mention the mixing of parts, and poor building practice of these unheard of manafactures, buying a power supply made by a different company than the light source itself! not to perform optimaly. All in all LED's r unpredictable, imo

I have used different LED types in various fixtures. They should be fine as long as you do not run them out of spec. A setup with components from different companies would not matter, as long as the specifications match. Matching specifications is more important than matching brand, as even with MH setups most people run bulbs made by a different manufacturer.

As for LED manufacturers, there are all types out there, just as is the case with other lighting manufacturers. There are several well known and respected LED fixture producers, as well as countless no-name chinese imports for sale. I am curious why you would say LEDs are not unpredictable? How do you come to that conclusion?

either one burns out and comprimises the rest in the same seriese or the controller goes, for the price of them.... í'll stick to the old school mh,hps and ho flores

If one burns out, replace it. One burning out would not compromise the rest of the series. It is like the old christmas lights, replace the one in the series that is out, and they all work again. It can be a relatively cheap fix at around $5-6 for a replacement LED.

Anyway, good luck in the future with reefing. I see from your signature you are new to the hobby. Might I suggest that you do a little more reading here on RC, as there is a wealth of information available. Don't be so quick to jump to conclusions about various technologies.

Landon

 

[Reffer831]

No!!!! They are great in the beginning for the first year or two then they dim out just like any other light.

 

[Anemonebuff]

CFLs and florescents in general are more difficult to keep at their optimal temp. Heat a CFL too much it will burn out and/or color shift quickly. Cool a CFL too much and it will not meet it's full potential. LEDs only enemy is heat. You can cool them to freezing temps and they will work just as good as a LED just at the max suggested temp. Just look at a CFL used outside in winter. Takes mine about an hour to heat up and give off the max light it can produce. It also takes the life out of the bulb more rapidly.

 

[Shard]

That is a good point about the CFLs Anemonebuff. I was only thinking in terms of LEDs and regulating the max temp, I forgot about CFLs having a narrower range both above and below for optimal output/life.

 

[Pink_n_Purple]

Yes i am very new to this hobby, and thanks Rc and members for the arcticles, and abundant info,awsome!

I was always under the assumsion that leds have a high life expectancy (longer than 2 years @12-12) thats just my oppinion, and multiple companies\manafactures claiming this.

Heat build up is there major down fall in life expectancy, which can result in current loss also. Most premade manafactures models dont coolo there systems well enough, they want yoyur leds to burn quicker, so you can buy another.

I'm sorry to say that most Diy are usually to cheap and or just take short cuts in try to creat there own heat sinks , that not cooled or at very best is a cheap attempt at doing so. It may be cheaper as a diy but replacement can be combersome due to poor planning and builds, also you have to make the decision of replacing them all or only a few a time and wich ones?(larger systems apply here) mutiple diodes increases chance for manafaturer mistakes also. shipping can be combersome and slow too if needed quicky.

IMO Led's should last you no shorter than 4 years at 12 on 12 off in other wise there just not worth the expense or trouble of replacing, and thats just my op.