Led: it is true that they have 50.000 hours of operation?

[frank88]

Hello everyone from Italy .

I'm about to order a razor maxpect 300 for my tank 48 ". My doubt is: I chose LED lighting because it is said that long-term savings is because consumption is less at the same efficiency. Otherwise I would have mounted a 'T5 lighting 6 or 8 * 54 * 54. I appeal to those who own or have owned a ceiling light LED: in fact we are sure that the duration of the LEDs is actually at least 50,000 hours, or after the time even the LEDs go meeting to wear losing intensity?
thanks

[wetWolger]

Specifically that "50,000" hours number is actually when LED's will lose 30% of their intensity. In addition what really kills LED's is heat, so that is assuming they are kept absolutely cool during operation (which is likely not 100% reality, even with a fixture that runs as cool as the maxspect razor does).

Running them at a lower intensity will help keep them cooler, and last longer. I have a maxspect razor as well, and will say mine runs VERY cool (not in a canopy or anything). But I still expect that if you were to drive it at 100% for a good chunk of the day you likely will have less than 50k hours before you reach that 30% loss (as those are ultra ideal conditions). How much less than 50k hours? That's hard to say. But I feel if it's even 1/5'th that (10k hours) I am doing pretty good (that would be around 3.4 years at 8 hours a day).

The real question is, does it matter? First if your not running at full intensity, that may mean you just need to increase your intensity over the years. But assuming you are both driving at 100%, and need to maintain that absolute maximum intensity....That is in part why I think it's important to get a fixture where the LED pads can be replaced without having to buy an entirely new fixture (which you can do with the maxspect).

[Dmorty217]

My buddy has been running the first DIY kit Rapidled ever came out with. Has only white and blue Leds and has been running them well past the 50,000 hr mark and his SPS grows like crazy. Also he has very little air flow to cool the Leds so I think the 50,000hr life span is way off the mark

[frank88]

Specifically that "50,000" hours number is actually when LED's will lose 30% of their intensity. In addition what really kills LED's is heat, so that is assuming they are kept absolutely cool during operation (which is likely not 100% reality, even with a fixture that runs as cool as the maxspect razor does).

Running them at a lower intensity will help keep them cooler, and last longer. I have a maxspect razor as well, and will say mine runs VERY cool (not in a canopy or anything). But I still expect that if you were to drive it at 100% for a good chunk of the day you likely will have less than 50k hours before you reach that 30% loss (as those are ultra ideal conditions). How much less than 50k hours? That's hard to say. But I feel if it's even 1/5'th that (10k hours) I am doing pretty good (that would be around 3.4 years at 8 hours a day).

The real question is, does it matter? First if your not running at full intensity, that may mean you just need to increase your intensity over the years. But assuming you are both driving at 100%, and need to maintain that absolute maximum intensity....That is in part why I think it's important to get a fixture where the LED pads can be replaced without having to buy an entirely new fixture (which you can do with the maxspect).
thanks for reply

yes, for me is matter. because I think that for now led light are not at the same level of the mh and t5 light
so, if it's really that they work for a long time, good.. otherwise I would prefer run T5.

[frank88]

My buddy has been running the first DIY kit Rapidled ever came out with. Has only white and blue Leds and has been running them well past the 50,000 hr mark and his SPS grows like crazy. Also he has very little air flow to cool the Leds so I think the 50,000hr life span is way off the mark

here in Italy we think (seeing corals colors and grow) that for now the best results remain using Metal halide or t5 light.

we have seen spectacular tank with these illumination, and normal/good tank with led


personally I think that perfect water parameters are the only way to maintain sps colour at max level, so or Mh, t5 or led for me are indifferent.

[Bpb]

I would be thrilled to see how many reefers used their leds for even half that amount of hours before fixing or upgrading their units. Rememeber folks, if you upgrade your units or even just your pucks every time a new generation comes out...how exactly is that any different from bulb replacement cost

[biecacka]

I was told by a rep that the 50k usage was in a controlled environment, no heat no humidity. We should expect about 3-5 years on average for our units with he wear and tear on them. Not sure but it is what I was told.....

Corey

[frank88]

I was told by a rep that the 50k usage was in a controlled environment, no heat no humidity. We should expect about 3-5 years on average for our units with he wear and tear on them. Not sure but it is what I was told.....

Corey
yes I asked an mechanical engineer this morning, it said me that max they work for 5 years..

[madweazl]

My buddy has been running the first DIY kit Rapidled ever came out with. Has only white and blue Leds and has been running them well past the 50,000 hr mark and his SPS grows like crazy. Also he has very little air flow to cool the Leds so I think the 50,000hr life span is way off the mark

If they were run 24 hours a day, 365 days a year, it would take 11.4 years to reach 50k hours. How long has he had those lights...

[frank88]

I was told by a rep that the 50k usage was in a controlled environment, no heat no humidity. We should expect about 3-5 years on average for our units with he wear and tear on them. Not sure but it is what I was told.....

Corey

If they were run 24 hours a day, 365 days a year, it would take 11.4 years to reach 50k hours. How long has he had those lights...

so how long you think they run at good intensity?

[ca1ore]

Specifically that "50,000" hours number is actually when LED's will lose 30% of their intensity.

This is the key point. Doesn't mean they will fail after 50K hours (they may fully fail before or after) but that they will have lost 30% of their intensity at that point on average. I'd actually argue that most of us wouldn't even notice. Plus, if you run your lights at something less than 100%, you have 'headroom' to compensate. Running your lights 12 hours/day, give you almost 11 1/2 years until 50,000 hours. I have certainly never kept a fixture for anywhere close to that length of time (at least, not in active use).

I don't think this is something to keep you awake at night :lol:

[power boat jim]

I ran my AI blues for 3.5 yeras at 10 hrs a day. I was lucky enough to be able to have a par meter available when I set them up and have one just before I took them down 3.5 years later. They showed no detectable loss in par over that period.

[frank88]

I ran my AI blues for 3.5 yeras at 10 hrs a day. I was lucky enough to be able to have a par meter available when I set them up and have one just before I took them down 3.5 years later. They showed no detectable loss in par over that period.

well, the first scientific experience
I'm very happy for this
a par meter is a only "judge" for this doubt

[power boat jim]

well, the first scientific experience
I'm very happy for this
a par meter is a only "judge" for this doubt

They were run pretty steady over the time I had them at 70/40/40 levels. I did the final measurement just to see what the degredation would be since at the time there was another thread about LEDs and how much they lose each year. I was surprised to see they lost nothing measurable on the PAR meter since that was not the general consensus. It is however only one case.

[frank88]

They were run pretty steady over the time I had them at 70/40/40 levels. I did the final measurement just to see what the degredation would be since at the time there was another thread about LEDs and how much they lose each year. I was surprised to see they lost nothing measurable on the PAR meter since that was not the general consensus. It is however only one case.


what led fixture you habe run? radion? vertex or what?

[power boat jim]

what led fixture you habe run? radion? vertex or what?

I started the hobby with MH. Changed to six AI blues to reduce the heat. They worked well, had decent growth and color. I added Full spectrum led strips to the tank with the AIs. This help some with color and growth. I recently sold the AIs and I am now running three Gen 3 Radion pros. I have had very good results with these. I need one more but since they cost a fortune, I stuck an old MH over the rear of the tank and run it a few hrs a day . The combination looks quite good and the resuts cant be argued with either.

[jedimasterben]

Just to clarify a few things in this thread. People are only getting bits and pieces of how LEDs are rated for their longevity.


Most LEDs today have an L70 (length of time until full intensity is reduced to 70% of starting output) rating of at least 30,000 hours, and I've seen some rated to nearly 100,000 hours. This is called lumen maintenance.

These lumen maintenance projections are made in very different conditions depending on the make and model of LED, but most are done in fixed conditions, usually tested at both lower and higher currents and temperatures, and in accordance with LM-80 test criteria according to the Illumination Engineering Society.

Let's take the Cree XT-E diodes. Here is the datasheet for how Cree determines lumen maintenance (http://www.cree.com/~/media/Files/Cree/LED%20Components%20and%20Modules/XLamp/XLamp%20Application%20Notes/XLamp_lumen_maintenance.pdf). Here are the results for the X-lamp series (http://www.cree.com/~/media/Files/Cree/LED%20Components%20and%20Modules/XLamp/XLamp%20Application%20Notes/LM80_Results.pdf). The XT-E is on page 40.

Run with 1A current (about 3.1 watts of power) and at 105C die and ambient temperature (this is enough to boil water at sea level and is far hotter than any aquarium fixture should get), 6,000 hours gives 96.1% lumen maintenance, leaving the L70 projection above 60,000 hours.

Pushing the diode to 1.25A (about 4.1 watts) with 85C die and ambient temperature, 6,000 hours gives 95.6% lumen maintenance, giving an L70 projection of greater than 54,000 hours.


Going to Philips for a minute, their Rebel ES and all later diodes are tested for their lumen maintenance at high currents and at 135 degrees Celsius and are rated for L70 at an average of 50,000 hours, so at lower currents or lower temperatures the L70 would be pretty significantly higher.


As for outright failures, you're going to get a few outliers throughout the realistic lifespan of the diodes, especially when you consider the multitudes of different solders used to attach to the PCB, the types of PCB used, the types of cooling used, etc. Realistically, though, failures will be few and far between going up to the L70 projection, and after that the failure rate begins to steadily rise as more hours are put on the diodes.

[jda]

Any of them say anything about spectrum shift over time, not just intensity? I mean, I have MH bulbs that probably COULD run for 8 years (about 30K hours), but who knows what they would look like.

[wetWolger]

I know that solid color (pure red, pure blue) LED's don't spectrum shift with age...they start very narrow and stay with it.

I actually don't know about white LED's since they produce their light via a completely different mechanism.

[jda]

They shift quite a bit when you dim them... like 20%ish IIRC, so they seem like they are plenty susceptible. Are you sure that somebody has published on this? Most everything that I have seen is illumination and intensity only.

[jedimasterben]

Any of them say anything about spectrum shift over time, not just intensity? I mean, I have MH bulbs that probably COULD run for 8 years (about 30K hours), but who knows what they would look like.
There is not much chromaticity shift if run properly (constant current driver, not too hot or too cold, etc).


They shift quite a bit when you dim them... like 20%ish IIRC, so they seem like they are plenty susceptible. Are you sure that somebody has published on this? Most everything that I have seen is illumination and intensity only.
Cree gives chromaticity shift in their lumen maintenance datasheet (linked above).

[slief]

If they were run 24 hours a day, 365 days a year, it would take 11.4 years to reach 50k hours. How long has he had those lights...

While I was thinking along similar lines, your math is flawed. 24 hours a day times 365 days is 8760 hours a year. That works out to about 5.7 years to reach 50,000 hours assuming 24/7 operation. That said, with a normal 12 hour photo period, you 11 years is about right.

As for the topic of this thread, I'm in the computer manufacturing business and have been for the last 20+ years. Diodes rarely ever fail so from the aspect of potential mean time before failure (MTBF) for an LED/Diode, the 50k power on hours is a fairly safe bet. That's not to say that there won't be some color shift or intensity loss during that period of time. The reality is that there are other components with the LED fixtures that are more likely to fail during that 50k hour period. Things like power supplies and driver boards have a host of other components such as capacitors, transistors and other stuff that is more likely to fail during that period of time than the diodes.

[frank88]

While I was thinking along similar lines, your math is flawed. 24 hours a day times 365 days is 8760 hours a year. That works out to about 5.7 years to reach 50,000 hours assuming 24/7 operation. That said, with a normal 12 hour photo period, you 11 years is about right.

As for the topic of this thread, I'm in the computer manufacturing business and have been for the last 20+ years. Diodes rarely ever fail so from the aspect of potential mean time before failure (MTBF) for an LED/Diode, the 50k power on hours is a fairly safe bet. That's not to say that there won't be some color shift or intensity loss during that period of time. The reality is that there are other components with the LED fixtures that are more likely to fail during that 50k hour period. Things like power supplies and driver boards have a host of other components such as capacitors, transistors and other stuff that is more likely to fail during that period of time than the diodes.then 50,000 of operation can be declared realistic?

[jedimasterben]

then 50,000 of operation can be declared realistic?
Depending on the diode, cooling, drive current, etc.

[frank88]

Depending on the diode, cooling, drive current, etc.
in a average situation

[alton]

The biggest issue with LEDS is the manufactures who are stuck on one year warranties. I use BML fixtures with a three year warranty. And as some of my friends who have found out when they lost multiple diodes after one year and had to pay $200 to upgrade it sux. In the commercial industry most manufactures are offering minimum 5 and maximum 10 year warranties.

[jedimasterben]

in a average situation
There are so many different types of LEDs and so many different cooling scenarios that there really is no 'average'. If the LEDs are kept reasonably cool (below 60C), not overdriven, and are of a quality make, then the LEDs should last that long or much longer.