LED lighting experiences.

[Furcifer]

trakar, sorry, have no interest in a sentence by sentence quoting argument. You haven't read my links. the Philips study I linked to purely tested the LED emitter, and that study is what they specify to support their lifetime claims. This is an LED emitter soldered to a test harness, not a consumer bulb. A description of the test, and a link to the report: http://www.philipslumileds.com/technology/lumenmaintenance. They clearly are not measuring and taking into account electronics failure. Something several other papers I linked to pointed out as a failure of the estimates by the manufacturers. I linked to a paper that specifically address the failure rates of the drive circuitry, which demonstrated that the lifetime is very difficult to model, and will generally be much less than the LED emitters themselves.


Sorry, but in the world I inhabit claims require evidence before I will believe them. I don't assume claims are true. If you have evidence that a consumer bulb has been tested to the claimed 22 years (for the Philips example), I'll listen, otherwise statements like "that's how capitalism works" is a nonstarter.

So, do you have evidence?

I'd point out again you're dealing with a new technology so you have to be prepared for the "new way of doing things so to speak". It isn't perfect yet and you're right to be skeptical. However.
You're dealing with times so long it's hard to accurately repeat in the lab. They've had to come up with ways of simulating 20 years. Obviously simulating 20 years on a radial tire is easier than on an LED so it can't be as reliable a measure. None the less the standards are there, they're based in science and they're universally accepted. I don't believe they've been created with the intention of misleading consumers, but I'm sure they will. I don't think there's any way of avoiding that.

Long story short; your demands are unreasonable given the technology, but perfectly understandable. In the real world we have to agree on some standard and run with this until we have 20 years of field testing to compare it to. In your world you don't.

(personally I'm confident in the technology, less confident the Chinese can build and assemble the components with the quality required to keep them running for +20 years.)

 

[Dan O.]

The failure modes of the LED bulbs would be quite well known since this is nothing more than a semiconductor junction and they've been studying those for decades. There are going to be three primary failures: high current through the bonding wire causes metal ion migration which can lead to weak spots in the wire. Thermal stress at the interface between the bonding wire and silicon chip will lead to cracking of the bond. And probably the most significant factor for high power LEDs, high temperature in the chip increases the rate that the trace elements migrate through the junction region all of these processes are well known and mathematically modeled so I don't doubt that the manufacturers would be able th estimate the expected lifetime of the light. Also, with the model, they can use accelerated lifetime testing by exaggerating those conditions that cause the failures.

 

[roger]

I'd point out again you're dealing with a new technology so you have to be prepared for the "new way of doing things so to speak". It isn't perfect yet and you're right to be skeptical. However.
You're dealing with times so long it's hard to accurately repeat in the lab. They've had to come up with ways of simulating 20 years. Obviously simulating 20 years on a radial tire is easier than on an LED so it can't be as reliable a measure. None the less the standards are there, they're based in science and they're universally accepted. I don't believe they've been created with the intention of misleading consumers, but I'm sure they will. I don't think there's any way of avoiding that.

Long story short; your demands are unreasonable given the technology, but perfectly understandable. In the real world we have to agree on some standard and run with this until we have 20 years of field testing to compare it to. In your world you don't.

(personally I'm confident in the technology, less confident the Chinese can build and assemble the components with the quality required to keep them running for +20 years.)

I understand that it is currently impossible to say what the true lifespan of modern bulbs is - heck, I linked to a paper explaining the science behind that difficulty. I'm just personally not willing to be the test subject, given the cost. I have no quibble with those forging ahead, so long as they understand the 20+ year estimates are merely lumen maintenance numbers.

I disagree about the standards. It has been widely admitted that the standards for LED lights are more or less made up. IES LM-80 is the standard the industry has settled on, but it is merely how long it takes a running LED to get to 80% brightness. We have no way to test/model actual lifespan, taking drivers, etc., into account. The science is not there yet. Yet manufactures blithely tout 20+ lifespans and use that to calculate energy savings. They know it is not a true lifespan estimate, yet they use it. Sounds like they are intentionally misleading customers to me.

Again, my situation. 17 recessed cans in this room alone. The Cree LR6 is a very well regarded light for that application. 80 bucks a pop from Amazon. Only 12 watts, which is still higher than recommended for recessed can use, but note the very large infrastructure around the bulb to provide cooling and ventilation. Nonetheless, it is 650 lumens, far below the 1700 lumens I'm getting from my halogens. I normally am running about 4 -8 cans at a time, with the reduced light levels, I'd probably be running most of the 17 KEDs at once, and have supplemental lamps going as well. For each light, I would save $9 a year iff I burn it 4 hours a day every day. But, I don't burn all 17, so I average maybe, what, 1 hour a day in the winter, and less than that in the summer. So, about a 40 year payback period? I should spend over $1300 dollars that I can recoup in 40 years, iff the bulbs actually last 3-4 times the warranted life span? For 1/3 the current light level? Nope. Not for me. Probably I'll end up with an LED in a lamp in the next year or two - that's a $20 experiment, so why not? Supposedly the well regarded Philips bulbs are coming out with higher lumens in the next few months, so it might be something I'd be willing to use.

Which is not to say it doesn't work for other people. It obviously does. And, I have no doubt, it'll eventually work for me. It's a great technology. It's just not there yet for many of us.

 

[roger]

For those disputing IES LM-80, and what it tests (Trakar, I'm looking at you ): http://apps1.eere.energy.gov/buildings/publications/pdfs/ssl/lightfair8_richman-ies.pdf

Go to page 10. This was created by the DOE. I quote:

Approved method for measuring lumen depreciation of solid-state (LED) light sources, arrays and modules
• Does not cover measurement of luminaires.
• Does not define or provide methods for estimation of life.

That's pretty cut and dried. Go to any bulb site - they quote their LM-80 results. LM-80 has nothing to do with bulb lifetime. Yet the #s are used everywhere to calculate costs.

 

[Dan O.]

By "used everywhere", are you referring to the places that use TM-21 to extrapolate a lifetime from the LM-80 test result?

 

[Furcifer]

I understand that it is currently impossible to say what the true lifespan of modern bulbs is - heck, I linked to a paper explaining the science behind that difficulty. I'm just personally not willing to be the test subject, given the cost. I have no quibble with those forging ahead, so long as they understand the 20+ year estimates are merely lumen maintenance numbers.

I disagree about the standards. It has been widely admitted that the standards for LED lights are more or less made up. IES LM-80 is the standard the industry has settled on, but it is merely how long it takes a running LED to get to 80% brightness. We have no way to test/model actual lifespan, taking drivers, etc., into account. The science is not there yet. Yet manufactures blithely tout 20+ lifespans and use that to calculate energy savings. They know it is not a true lifespan estimate, yet they use it. Sounds like they are intentionally misleading customers to me.

Again, my situation. 17 recessed cans in this room alone. The Cree LR6 is a very well regarded light for that application. 80 bucks a pop from Amazon. Only 12 watts, which is still higher than recommended for recessed can use, but note the very large infrastructure around the bulb to provide cooling and ventilation. Nonetheless, it is 650 lumens, far below the 1700 lumens I'm getting from my halogens. I normally am running about 4 -8 cans at a time, with the reduced light levels, I'd probably be running most of the 17 KEDs at once, and have supplemental lamps going as well. For each light, I would save $9 a year iff I burn it 4 hours a day every day. But, I don't burn all 17, so I average maybe, what, 1 hour a day in the winter, and less than that in the summer. So, about a 40 year payback period? I should spend over $1300 dollars that I can recoup in 40 years, iff the bulbs actually last 3-4 times the warranted life span? For 1/3 the current light level? Nope. Not for me.

Which is not to say it doesn't work for other people. It obviously does. And, I have no doubt, it'll eventually work for me. It's a great technology. It's just not there yet for many of us.

You have to remember this technology is basically backwards from a traditional light bulb in terms of rating; an incandescent or CFL is basically dead when you buy it, it lasts as long as it lasts. LED's last forever and then you subtract from that. (kinda like a car I guess)

As for your room, if you were serious about LED's for a room that large I believe you would want a DC unit, LED's, reflectors and drivers. It would be fairly simple but would take some time and planning. In the end it would save you considerable money over the "store bought" solution we're talking about here. Obviously if you were just building the house this would be much more practical as you wouldn't have put out the money for fixtures which aren't necessary.

That might not be practical for you though, maybe you don't have the time or the tools or the knowledge to do so. I don't know if you could find an electrician doing this yet, although you really don't need one (IANAE, check codes in your area)

Anyhow, not to lecture you, but I think it's wrong to say the technology isn't ready just yet. The technology is really there, it's just not packaged for mass consumption. We should start to see more custom jobs in bars and restaurants, even homes and businesses that will change how we view lights. I'm not sure we'll even want them to resemble lights like we have them now.

 

[mhaze]

It is not my claim, it is what the manufacturers of the bulbs say. I linked to one that specifically said to limit use to 10watt bulbs in recessed ceiling lights.
Googling reveals much more of the same. Why am I being ridiculous to listen to manufacturers? Would you care to speculate that when you google LED recessed lighting, you get endless hits for "retrofits" - meaning you get a new can and an LED lightbulb, and that these retrofit cans are designed for ventilation?

I guess that because I've done quite a few heat transfer problems, using computer modeling, traditional paper and pencil and BTU transfer rates, and intuition, I have never bothered to "look things up on Google". Incidentally this work was both with heating and airconditioning, engine cooling, and semiconductor cooling. Issues such as heat transfer in a ceiling can just seemed second nature and obvious. I already knew about degraded lifespans of components under heat stress.

Of course, that doesn't mean that I'm always right or don't over look things - like I never considered lightning strikes as being an issue with LED bulbs. Of course they're not grounded, which helps a lot.

Regarding your question about ventilation, old cans are quite leaky. They are rated as qualified for being buried in ceiling insulation or not (NOT means possible fire hazard). Thermal switch prevents over temperature operation in the dangerous range. Today with people concerned more with energy costs, it is considered "good" to have a sealed can with no air leakage. In such a fixture, since hot air rises, some accumulation of heat would be expected in the can based on outflow (radiant) against waste heat produced. Essentially this partly or fully negates the conductive utility of the finned radiator of the bulb.

So there isn't ONE ANSWER regarding ceiling cans - just as for putting even incandescent bulbs in a ceiling can, there isn't "one answer". Same true for CFL. The specific can, the insulation above, and the bulb's heat must be mutually considered. Not complicated, but not "one answer".

By the way, those 8 bulbs that have been running for 43K hours here survived a lightning strike about 75 feet away that blew out the pole mounted transformer. Doesn't really mean much, I guess.

 

[mhaze]

....if you were serious about LED's for a room that large I believe you would want a DC unit, LED's, reflectors and drivers.....The technology is really there, it's just not packaged for mass consumption. We should start to see more custom jobs in bars and restaurants, even homes and businesses that will change how we view lights. I'm not sure we'll even want them to resemble lights like we have them now.

Just for grins take a look at this - it has a medium edison base, but runs on DC. So in fact you could simply take a home AC circuit, and connect it to a DC source, screw some of these bulbs in, and eliminate the AC-DC conversion issue a the bulb level.

http://store.marinebeam.com/e26-e27-12v-24v-screw-base-led.html

 

[BenBurch]

Furcifer, Basically correct, but the technology is still rapidly advancing. The lumens/watt over date graph is very steep right now, and there are probably a few years before it levels off.

 

[Furcifer]

Just for grins take a look at this - it has a medium edison base, but runs on DC. So in fact you could simply take a home AC circuit, and connect it to a DC source, screw some of these bulbs in, and eliminate the AC-DC conversion issue a the bulb level.

http://store.marinebeam.com/e26-e27-12v-24v-screw-base-led.html

Interesting option. This is typical of the niche market stuff you see from smaller manufacturers taking advantage of LED's.

Curious, the base looks like a CFL base. I bet you could recycle CFL's buy disposing of the nasty chemicals inside and turning them into one of these. If you were so inclined.

 

[Trakar]

For those disputing IES LM-80, and what it tests (Trakar, I'm looking at you ): http://apps1.eere.energy.gov/buildings/publications/pdfs/ssl/lightfair8_richman-ies.pdf

Go to page 10. This was created by the DOE. I quote:


That's pretty cut and dried. Go to any bulb site - they quote their LM-80 results. LM-80 has nothing to do with bulb lifetime. Yet the #s are used everywhere to calculate costs.

This states that the IES (Illuminating Engineering Society of North America) document LM-80-08 ("Approved Method: Measuring Lumen Maintenance of LED Light Sources - 2008) , "Does not define or provide methods for estimation of life."

What I stated, is that the DoE guide document I linked to, published in June of 2011 (LED LUMINAIRE LIFETIME: Recommendations for Testing and Reporting, reconfirms all of the very problems and issues of concern we have all acknowledged, and then makes recommendations and and addressments for how to account for these issues moving forward.

Regardless, however, the issue of primary concern to consumers isn't that LED lamps exhibit a strong tendency to cease functioning at a shorter time period than is implied by their LM₇₀ rating in the IES LM-80-08, but rather that in typical application LED lamps tend to have a useful lifespan much greater than the industry is promoting for its products. I agree that better clarification of these issues is important, and apparently everyone in the industry agrees as well, but there is nothing in the currently available data which suggests that the current confusions result in an overestimation of LED bulb lifetime.

 

[Trakar]

trakar, sorry, have no interest in a sentence by sentence quoting argument. You haven't read my links. the Philips study I linked to purely tested the LED emitter, and that study is what they specify to support their lifetime claims. This is an LED emitter soldered to a test harness, not a consumer bulb. A description of the test, and a link to the report: http://www.philipslumileds.com/technology/lumenmaintenance.

I have read every link you provided when communicating with me, and sought links you provided to others. This (above) link is not to a study, but to a corporate web page reporting some graphs from a study* which apparently examines LED emitter testings. (BTW - this Phillips page states "Lumen maintenance and lifetime are intertwined," in their assessments.
(*LM-80 test report for the LUXEON Rebel -http://www.philipslumileds.com/uploads/294/DR04-pdf)

They clearly are not measuring and taking into account electronics failure.

That may be because you are looking at the LM-80 test data (lumen maintenance testing) instead of the LM-79 test data (Electrical and Photometric testing). As repeatedly stated, however, there are ways to incorporate both of these data sets into an overall lifetime rating as recommended by DoE (and generally included in newer Energy-Star testing and rating regimens).

Something several other papers I linked to pointed out as a failure of the estimates by the manufacturers. I linked to a paper that specifically address the failure rates of the drive circuitry, which demonstrated that the lifetime is very difficult to model, and will generally be much less than the LED emitters themselves.

I do not recall this paper but would be happy to examine it. Given, however, that the emitters themselves may continue output under ideal conditions for several hundred (or more) hours, it is not surprising that the electronics, especially in suspect manufacturing circumstances, may be the weaker link. This does not mean that the electronics are failing sooner than the LM₇₀ est., and I am unaware of any reports or studies that suggest that such premature failures are occurring at anything more than a statistically unusual rate. If you have data which compellingly indicates that such is the case, please present it.

Sorry, but in the world I inhabit claims require evidence before I will believe them.

Please present the evidence that indicates to you that LED lamps in typical application are failing at rates higher than industry LM₇₀ rating warantee standards.

So, do you have evidence?

You are the one claiming/fearing that the actual lifespan is significantly less than the LM₇₀ rating, I await your supporting evidences.

 

[Alferd_Packer]

I live on a boat and thought that LED's would be teriffic. Reality rears it's ugly head again. 12V LED's are still quite expensive ($20.00 range) and not very good. Reading lights don't work near as well as traditional's. Don't know if it's my imagination but LED light doesn't seem to carry very far either. I do find the "blue" to be annoying as well. Flourescent's seem to be much better light and draw minimal amps.

I looked into LEDs for my popup camper last fall. The bulbs are somewhat expensive, but I only have 4 of them.

 

[mhaze]

Interesting option. This is typical of the niche market stuff you see from smaller manufacturers taking advantage of LED's.

Curious, the base looks like a CFL base. I bet you could recycle CFL's buy disposing of the nasty chemicals inside and turning them into one of these. If you were so inclined.

No, the base here would serve little use as far as I can see. The LED requires 3-6v, the unit operates on 12-24v. Both are DC. the CFL electronics has to jack the 110v way up to drive the fluorescent.

In other words, if you ran 6 volts in your house, you could switch CREE led components directly on and off. You'd have a lifetime for them based on strictly the LED component. There would be no additional electronics.

 

[Furcifer]

No, the base here would serve little use as far as I can see. The LED requires 3-6v, the unit operates on 12-24v. Both are DC. the CFL electronics has to jack the 110v way up to drive the fluorescent.

In other words, if you ran 6 volts in your house, you could switch CREE led components directly on and off. You'd have a lifetime for them based on strictly the LED component. There would be no additional electronics.

Oh, I meant dropping a driver in the bottom. I've got a whack of 3-18V drivers that I got for a buck a piece. I bet they would work even with 24V but that's probably pushing it.

Is 6V in the range? I haven't looked at the specs lately. A 6V direct drive system would probably be the best, although I haven't seen a 15 or 30A AC/DC converter (I'm sure they exist though).

 

[mhaze]

Ah, that approach. You would get a kick out of this website.

www.superbrightLEDS.com

Caution: It would take DAYS to go through all the goodies (and I suspect it grows monthly).