LED lighting experiences.

[roger]

Here is an interesting article on the true lifetime of LED consumer bulbs.

It's a couple years old, so I welcome newer data. However, the point is essentially what i have been making: the critical component is not the LED emitter, but the drive circuit, and no manufacturer is calculating MTTF for the drive circuits.

The life of an LED driver is mainly determined by the lifetime of the electrolytic capacitors employed. Therefore, to achieve long life of the LED drivers, it is critical to select long-life, quality electrolytic capacitors (as discussed in March 2009 article). It is important to note that the life of electrolytic capacitors drops by 50% for every 10C increase in operating temperature and so thermal management of these components is extremely important.

As stated in the March 2009 article there are several Achilles’ heals within LED drivers including electrolytic capacitors and components such as MOSFETs that can run close to their maximum temperature specifications however it is currently difficult to define the lifetime of an LED driver.

There are several methods for defining reliability that use very sophisticated mathematics and statistics however in practice it is common to find products that fail either straight away or after a significant period of use and this distribution is commonly referred to as a bathtub distribution. Figure 8 shows the reliability ‘bathtub curve’ which models the cradle to grave instantaneous failure rates vs. time.

I would like to leave you this month with the thought of how much work still has to be done within the LED industry in order to raise quality. I was recently at an event held that matched LED companies and end users from the public sector together and one of the companies told me they could buy a 20 or 30W LED driver for less than $5 from China and their aspirations were to manufacture an LED downlight for less than $20 complete within 12 months. After asking several questions about quality and sustainability it became very clear that these were not at the top of this company’s agenda which is of great concern.

I predict there is a real danger for the reputation of the LED industry and hence the wider lighting community when end users see LED lighting products that claimed 50,000 hours life start to fail in a few years time. ]

No one really knows, yet, how these new bulbs will perform. I'm not sure where the super cheap bulbs are going to come from, however, given the reliability depends so heavily on things like capacitors and heat management, and capacitors are a mature technology where long life in heated environment still costs a lot of money.

 

[roger]

Most of the LEDs I am buying do not promise 50K hours. Typically they say 10K hours, sometimes 20K. Being an "early adopter" I have a group of 8 installed since 2006. One has failed, the others are somewhat more yellow than when installed. That's running non stop for 5 years, incidentally, because that group, we don't bother to turn off

Not turning it off helps you, as the driver is not experiencing thermal changes which shorten the lives of capacitors and resistors. Philips claims 22.8 years, which is 33Khours assuming 4 hours/day usage (source) And, again, the numbers are NOT MTTF, it's a meaningless comparison.

RE heat accumulatation and dissipation: The solid state control circuit in an LED needs only to produce low voltage DC, while that in a CFL is considerably more complex. The same issues you mention are more true with CFLs. But no one ever mentions that, do they?

At $5 vs $50 it's less important. But yes, go to amazon and read CFL reviews, and you will find plenty of discussion of the need for good air flow, etc., to avoid early failure.

But, I don't see the point in worrying about what people are saying about CFLs. The reality is that LEDs are susceptible to heat and driver design. Solving both is currently expensive.

 

[Furcifer]

The harder you make it, the less you are really standing behind your product.

I'm not so sure about that. I had an RWS rifle that had lifetime service but I had to keep the box to return it in. Likewise all of my Benchmade knives have to be returned in the original box.

I'm pretty sure it's about protecting themselves from dishonest customers, not making it hard for the original purchaser for whom the warranty is for.

The few times I've had to return something to the manufacturer without the box or receipt they've been good about accepting the item once.

 

[roger]

Why do you hate the Earth?

**** the earth.

 

[roger]

I'm not so sure about that. I had an RWS rifle that had lifetime service but I had to keep the box to return it in. Likewise all of my Benchmade knives have to be returned in the original box.

I'm pretty sure it's about protecting themselves from dishonest customers, not making it hard for the original purchaser for whom the warranty is for.

The few times I've had to return something to the manufacturer without the box or receipt they've been good about accepting the item once.

Hmm, are we going to see light bulbs being sold on craigslist and such, such that this would be a concern? I guess we could, if they bulbs live up to the hours projections, and prices don't come down.

edit: In just two rooms I counted 62 bulbs. So, around 150 for the entire house. Pretty ridiculous to keep that many boxes & receipts around. Do you have the carton to your fridge, stove, garbage disposal, fans, kitchen aid, toaster, microwave, speakers, tv, stereo, electric toothbrush, shavers, etc? I don't. My mother did, at least for the smaller stuff but she was rather OCD. Companies know very well that the more conditions you place on returns, the less likely they are to receive returns.

 

[Trakar]

I was responding to the claim about not needing ventilation. The opposite is true. While LEDs do not put much heat out compared to other light sources, they are far more sensitive to that heat, and heat dissipation is a huge concern if you don't want premature failure. Stick that $40 bulb in a recessed ceiling? Don't plan on it lasting for long.

I'd be interested in reading some real customer reviews, from sources where there isn't likely to be much politically motivated spoofing.

As for reliability vs cheapness, sure, cheap imports won't last long. OTOH, we have plenty of reports, and I have seen, LED traffic lights going out.

again, I'd be interested in reading those reports. In the case of traffic lights, and even automobile brake lights, I would think that the constant cycling and environmental exposure, and possibly even the character of the power supplied, would be much larger issues of concern to lifespan than in most other applications.

The claims of 50,000 hours are entirely specious...

specious? really? would you object to me calling your usage of this qualification entirely specious?

- all that number means is that it takes 50,000 hours to go to 70% of light output due to dimming. It says nothing about the MTTF of the bulb, of the electronics, of damage and color shifts of the lens, etc.

No, but neither do anecdotal reports of failures or extraordinary longevity beyond manufacturer claims.

The reality is that these things are failing.

at a rate above expected? got link to statistical analyses or reliable, evidence based information supporting this assertion?

 

[roger]

I'd be interested in reading some real customer reviews, from sources where there isn't likely to be much politically motivated spoofing.

Why would you want that, when the manufacturers say they are sensitive to heat, and not to use them in these applications, or to severely downrate them if you do? I provided links, and google provides many more.

specious? really? would you object to me calling your usage of this qualification entirely specious?

What you say does not reflect on me, so go ahead. The fact remains you cannot compare MTTF numbers with dimming numbers.

at a rate above expected? got link to statistical analyses or reliable, evidence based information supporting this assertion?

I have no interest in trying to support something I did not claim or assert. The claim is that we (where we may = I, as I welcome data) do not have MTTF data on LED consumer bulbs. As I said in another post, it might turn out that the MTTF is far in excess of the 25Khrs (or whatever) that is claimed. We just don't know. But, like I said, we know they are failing, and that at least some manufactures are using cheap drivers with corresponding short life times. I've seen no data on the apparently good new Philips bulbs. So, we don't know with those.

So, got any data on MTTF of current generation consumer LED bulbs? I would genuinely welcome them. I'd like nothing better than to reduce my electric bill, as I spend my evenings in relative darkness due to cash flow problems.

edit: I neglected to answer the traffic light question. I agree it is apples/oranges. Nonetheless, it is interesting when they do not perform to claimed standards, yes. Here is a survey on LED usage in CA. I'll note only 56% of those surveyed said that it reduced maintenance costs, and 37% said that 12% of their lights burned out early. However, let me point out - that 37% is the percentage of people that responded to the question. i.e. 37% responded, and on average 12% failed early. That does not mean that 63% of those surveyed did not have early failures: we have no data on that.

Again, my only claim vis-a-vis MTTF is that we don't know the answer - we just have anecdotes. I further suggest that what we know about electrolytic capacitors is that making a cheap drive isn't really feasible with current technology.

edit: I guess it is annoying to just be given a PDF link. Here is a summary of some of the comments they received. Right above it is an interesting graph that I don't know how to copy and paste. One of the things is that out of 38 respondents, only 3 reported that the manufacturer honored the warranty. That squares with obviously suspect claims you can google about manufacturers not honoring warranties. I also recognize this was 2005. More current studies would be great so we could better track progress (I have no doubt progress is being made)

Some example of problems and comments cited include:
• The consistent failure of bulbs led to unusual patterns in the signal lens.
• LED failure was due to power supply problems.
• LEDs have not reduced maintenance cost.
• Maintenance costs remain the same but can increase as LED’s go out of warranty. More costly to replace than non-LED bulbs.
• The first batch of LEDs had problems.
• We experienced as many lamp outages with the LEDs as we did with the incandescents.
• Estimated age and color of LEDs needing replacement: varies with type of LED, early types do not last as long. A large purchase and installation in a short period guarantees a large purchase and installation of replacement LEDs in the future. However, early replacement of about 20 percent each year, for five years, spreads the cost.
• On going LED replacement on a four year cycle.
• LED replacement plan of 15 intersections per fiscal year.
• We plan to replace our reds in three years; all others in five years.
• The installation occurred over a three year period to avoid a mass replacement in the future. The LEDs are a great benefit to the city in terms of energy use reduction and maintenance cost.

 

[Furcifer]

Hmm, are we going to see light bulbs being sold on craigslist and such, such that this would be a concern? I guess we could, if they bulbs live up to the hours projections, and prices don't come down.

edit: In just two rooms I counted 62 bulbs. So, around 150 for the entire house. Pretty ridiculous to keep that many boxes & receipts around. Do you have the carton to your fridge, stove, garbage disposal, fans, kitchen aid, toaster, microwave, speakers, tv, stereo, electric toothbrush, shavers, etc? I don't. My mother did, at least for the smaller stuff but she was rather OCD. Companies know very well that the more conditions you place on returns, the less likely they are to receive returns.

Get off my lawn! hehe

Different times require different measures. I really don't expect this to last for very long. Like you pointed out, it's the drivers not the bulbs that are the issue and in the not too distant future you'll see LED's with a slot to pop in a new driver. The rest of the "bulb" will undoubtedly last "forever", it's plastic and a bit of metal.

 

[roger]

Get off my lawn! hehe

Different times require different measures. I really don't expect this to last for very long. Like you pointed out, it's the drivers not the bulbs that are the issue and in the not too distant future you'll see LED's with a slot to pop in a new driver. The rest of the "bulb" will undoubtedly last "forever", it's plastic and a bit of metal.

That's brilliant, didn't think of replacing only the driver. And, longer term they are talking about rewiring houses for DC, which should (if I understand it correctly) reduce or remove the driver issue, at which point the truly large LED emitter MTTF become significant.

 

[BenBurch]

**** the earth.

Use a condom.

 

[Trakar]

Why would you want that, when the manufacturers say they are sensitive to heat, and not to use them in these applications, or to severely downrate them if you do? I provided links, and google provides many more.

The only link I saw from you was an editorial written a couple years back, seemingly recounting data pulled from several manufacturer and testing sites dating back several years previous to the article's pub-date. None of the actual related data, however, supports that LEDs are failing at an unusually high rate out of proportion with manufacturer claims and advertising,...at least not that I was able to discern. The author does express his own concerns and fears, but he doesn't offer any evidence that his fears had come to fruition.

I will go back and look at earlier posts and see if there are links that I missed.

The fact remains you cannot compare MTTF numbers with dimming numbers.

without any indication that MTBF numbers are higher than dimming numbers, what difference does this make?

So, got any data on MTTF of current generation consumer LED bulbs? I would genuinely welcome them. I'd like nothing better than to reduce my electric bill, as I spend my evenings in relative darkness due to cash flow problems.

With 65 lamps in 2 rooms, it is easy to understand your interest. I'll give it a look see over the next few days and see what turns up.

edit: I neglected to answer the traffic light question. I agree it is not apples/oranges. Nonetheless, it is interesting when they do not perform to claimed standards, yes. Here is a survey on LED usage in CA. I'll note only 56% of those surveyed said that it reduced maintenance costs, and 37% said that 12% of their lights burned out early. However, let me point out - that 37% is the percentage of people that responded to the question. i.e. 37% responded, and on average 12% failed early. That does not mean that 63% of those surveyed did not have failures: we have no data on that.

Again, my only claim vis-a-vis MTTF is that we don't know the answer - we just have anecdotes. I further suggest that what we know about electrolytic capacitors is that making a cheap drive isn't really feasible with current technology.

(that is a 7 year old survey dealing in most cases with early generation LED lights that were installed several years prior to the survey.)

Not having data of extraordinary failure rates after a decade or so of extensive testing and fairly widespread application doesn't strike me as validating the impression that we should avoid the technology because there might someday be data indicating a higher than expected failure rate.

 

[mhaze]

That's brilliant, didn't think of replacing only the driver. And, longer term they are talking about rewiring houses for DC, which should (if I understand it correctly) reduce or remove the driver issue, at which point the truly large LED emitter MTTF become significant.

In many cases, there are incentives to run DC anyway. It doesn't require conduit or building permits. Further, cat5e computer cable POE is becoming more widespread and can easily support these loads.

In several existing application areas, LED 12v is a direct replacement for incandescent. Example is the small halogen decorative lighting me16 type bulb.

 

[roger]

The only link I saw from you was an editorial written a couple years back,<snip>

I will go back and look at earlier posts and see if there are links that I missed.

post 59. But you dispute that LED lamps have heat issues when placed in various closed enclosures?

without any indication that MTBF numbers are higher than dimming numbers, what difference does this make?

I don't understand your question. But I am referring to this article, which I linked to above, and many like it which I didn't link to.

Typical incandescent bulbs last 1,000 to 2,000 hours. But in speaking about LED replacements, lamp life is routinely quoted as 25,000 to 50,000 hours. Long lamp life, and the reduced power used to create the same amount of light, is what makes this technology so promising.
But what does a 25,000-hour life mean? As it turns out, no one is quite sure yet. The definitions surrounding LED lamps, a nascent technology, are still being made up as we go along.

Heck, here's a bunch more links, all touting really long lifespans. But, as I have said several times (and so did you, up thread, these numbers are not MTTF, just an estimate of how long it takes the bulb to dim).
http://www.electronicsweekly.com/blogs/led-lights/2009/02/led-life-expectancy.html
http://gizmodo.com/5151865/led-bulb-life-spans-are-not-what-they-seem
http://www.lunaraccents.com/educational-white-LED-life.html
http://techcrunch.com/2009/02/11/confusion-surrounds-led-light-bulb-lifetime-ratings/


And on and on. Heck, Carnegie Mellon assumed 25,000 hours was the lifetime of the bulb (http://bits.blogs.nytimes.com/2009/02/13/do-energy-saving-led-lamps-save-energy/), but as we know, that is not at all proven.



The point is that on incandescent packaging, they print 1000 hours (or whatever), and on LEDs they print 25,000hours (or whatever), but they do not mean the same thing. People in this thread and other threads are doing calculation like "okay, so the LED lasts 25 (25000/1000=25) times longer, so the cost savings are....". Except, we have about zero data on how long LED bulbs really last. We certainly know early ones did not last very long at all, and I linked a 2009 article saying the problem persists. Which leads us to...

Not having data of extraordinary failure rates after a decade or so of extensive testing and fairly widespread application doesn't strike me as validating the impression that we should avoid the technology because there might someday be data indicating a higher than expected failure rate.

Well, we have clear evidence that the industry has in the past supplied us with bulbs with claims that could not be met. At least now when you go to, say, Philips, they talk about lumens maintenance, not lifetime, if you go to their technical section. Of course, the bulb packaging still just says "lasts 22 years" or whatever. Personally, this leads me to not trust them. As I've said, I have hopes, because certainly the new Philips bulbs are getting rave reviews, and I have not seen one report of a failure, but who knows? But, personally, I'll let early adopters take on that risk. I don't see any compelling reason to think that in 2005 they said 25,000 hrs and it meant 3 weeks, but now when they say 22 years it really means 22 years. Again, data is totally lacking, and I linked to a paper that explained that it is in fact extraordinarily hard to model this accurately. So, we wait and see.


edit: ah, the duplicity continues. Here is a quote from a Philips spokesman, who claims the 25,000hr figure for their new bulbs, comparing it to the 5000MTTF of incandescents: "An individual EnduraLED, which will last at least 25,000 hours and probably even longer, will save consumers and businesses $120 in electricity costs over its lifetime" (source). It's bull. You can't compare MTTF to dimming time. It makes no sense.

And, while I'm editing, don't get me started on bulbs that emit 300-400 lumens being advertised as 60 watt replacements. 850 is a 60 watt replacement, though I'll concede 800 which is what several manufacturer use as it is the energy start standard. The advertising is bull in many cases, and my view is rather jaundiced at this point.

 

[Trakar]

In many cases, there are incentives to run DC anyway. It doesn't require conduit or building permits. Further, cat5e computer cable POE is becoming more widespread and can easily support these loads.

In several existing application areas, LED 12v is a direct replacement for incandescent. Example is the small halogen decorative lighting me16 type bulb.

It would certainly make sense as we (hopefully) shift to a more distributed, point/locality-of-use, power generation systems. Still need inverters to handle excess grid sales and supplemental grid purchases but as long as there is appropriate standardization it shouldn't be that big of an issue, especially if end-users are reconverting the power to DC prior to sending it to our home electronics.

 

[Furcifer]

The thing is LED technology doesn't have any problems. The problem is your home is wired for AC and your fixtures for a 130 year old design.

 

[mhaze]

It would certainly make sense as we (hopefully) shift to a more distributed, point/locality-of-use, power generation systems. Still need inverters to handle excess grid sales and supplemental grid purchases but as long as there is appropriate standardization it shouldn't be that big of an issue, especially if end-users are reconverting the power to DC prior to sending it to our home electronics.

Well, don't fall into a logical fallacy we've been seeing with CFL, namely the sweeping generalization of "it's good" or "it's bad". We've already got low voltage DC commonly used for doorbells, outside lighting, etc, and we know that cat5e can be run all over to supply low voltage DC to electronics, cams, in the order of 25-50 watts per line. All these have their place, as does ac wiring. ''

Roger's argument against LED is really an argument against supplemental electronics to run the LED from ac wiring, not an argument against LED at all.
Buy hey, he acknowledged that and his comments are more or less valid with respect to off the shelf LED units to an extent: As I noted, most of the stuff I just installed was listed on the box at 10K hours, which I think is conservative, and I've not seen any that was stated at 25K or 50K.

The stuff I put in in 2006 has ran nonstop since then, 5 years x 24 x 360 = 43,200 hours. One failure of about 8 or 10 total, and that was likely mechanical. (they were not even rated for outdoors, these bulbs). 2006 was ridiculously primitive compared to what we are getting now. Read into that whatever you like, but that's what I've got for data.

I could go take my point and shoot thermometer and provide data as to temperature in the interior of a surface mount ceiling can with a 40 and a 65 watt LED, but I know without doing this that the temps in there don't get up to levels that would seriously affect electronics. But the argument against LED in a ceiling can based on heat accumulation seems to be an argument in the direction of "the heat sink doesn't work or isn't engineered properly", which is a bit ridiculous.

Ceiling cans are drafty by design, since they had to dissipate heat (say 10x the amount from LED) from incandescent. Yes in that mode the temp rise in the can was permissible but those cans also have thermister shut down. Given this, Roger's argument is actually that the thermal switch is improperly set for LED maximal lifespan. I'd agree with that in principle, there are some other issues such as "IC and non-IC" and "sealed vs. non sealed" but they are not primary drivers.

 

[Trakar]

post 59. But you dispute that LED lamps have heat issues when placed in various closed enclosures?

it is certainly an issue for consideration and potential improvement (especially in older lamp designs), but in typical application today does not seem to be a major design problem nor premature failure concern. With regards to those links you might prefer/like:

http://apps1.eere.energy.gov/buildi...ssl/led_luminaire-lifetime-guide_june2011.pdf

I don't understand your question. But I am referring to this article, which I linked to above, and many like it which I didn't link to. Heck, here's a bunch more links, all touting really long lifespans. But, as I have said several times (and so did you, up thread, these numbers are not MTTF, just an estimate of how long it takes the bulb to dim).

If an LED lamp's electronics short out the first time it is turned on, is it dimmer than 70% of its total output? What leads you to conclude that investigations and testing of LEDs does not include these premature failures in with the bulbs that have merely dimmed by 70% when they are calculating failure rates? (at least DOE assessments seem to as mentioned in link above)


http://www.electronicsweekly.com/blogs/led-lights/2009/02/led-life-expectancy.html

Mentions considerations, no mention that any or all of these facotrs are resulting in any kind of significant, outside of manufacturer claims, premature failure rates in LED lamps.

http://gizmodo.com/5151865/led-bulb-life-spans-are-not-what-they-seem

A blog article restating the NYT editorial

http://www.lunaraccents.com/educational-white-LED-life.html

lot's of legitimate concerns mentioned, no data offered supporting that any of these factors are responsible for a greater than expected and advertised failure rate among general LED application over the last decade or so.

http://techcrunch.com/2009/02/11/confusion-surrounds-led-light-bulb-lifetime-ratings/

About three years old and primarily derived from the same apparent news release, again no data compellingly indicating that LED lamps were failing at an abnormally high rate

And on and on. Heck, Carnegie Mellon assumed 25,000 hours was the lifetime of the bulb (http://bits.blogs.nytimes.com/2009/02/13/do-energy-saving-led-lamps-save-energy/), but as we know, that is not at all proven.

you are correct the LEDs are probably good for a million hours and under the proper circumstances perhaps longer.

The point is that on incandescent packaging, they print 1000 hours (or whatever), and on LEDs they print 25,000hours (or whatever), but they do not mean the same thing. People in this thread and other threads are doing calculation like "okay, so the LED lasts 25 (25000/1000=25) times longer, so the cost savings are....". Except, we have about zero data on how long LED bulbs really last. We certainly know early ones did not last very long at all, and I linked a 2009 article saying the problem persists. Which leads us to...

You are probably correct the ultimate price savings for most individuals will probably be much greater than is typically estimated.

Well, we have clear evidence that the industry has in the past supplied us with bulbs with claims that could not be met. At least now when you go to, say, Philips, they talk about lumens maintenance, not lifetime, if you go to their technical section. Of course, the bulb packaging still just says "lasts 22 years" or whatever. Personally, this leads me to not trust them.

Agreed it aggravates me when people sell me products that are most likely going to perform better, longer and save me more money than they asserted.

As I've said, I have hopes, because certainly the new Philips bulbs are getting rave reviews, and I have not seen one report of a failure, but who knows? But, personally, I'll let early adopters take on that risk. I don't see any compelling reason to think that in 2005 they said 25,000 hrs and it meant 3 weeks,..

any evidence that this occurred at statistically significant levels,...or is this mere rhetorical hyperbole?

...but now when they say 22 years it really means 22 years,...

or based on some industry considerations, perhaps much longer.

Again, data is totally lacking, and I linked to a paper that explained that it is in fact extraordinarily hard to model this accurately. So, we wait and see.

The data isn't lacking, it's just seems not to be the type of data you wish to accept as compelling.

edit: ah, the duplicity continues. Here is a quote from a Philips spokesman, who claims the 25,000hr figure for their new bulbs, comparing it to the 5000MTTF of incandescents: "An individual EnduraLED, which will last at least 25,000 hours and probably even longer, will save consumers and businesses $120 in electricity costs over its lifetime" (source). It's bull. You can't compare MTTF to dimming time. It makes no sense.

I don't see the duplicity. Industry claims like these generally aren't about technical details as much as they are about product marketing and how long manufacturers want to carry warrantees on their products. If you feel that these statements are grossly out of line with the typical home user application of lamps, make your case and present your supporting evidences.
I suppose the intellectually rigorous thing to do would be to write to the Phillips spokesman and ask him to present his supporting evidence or retract his statement. I've simply seen no rising tide of anecdotal or circumstantial evidence suggesting that these products are unusually prone to premature failure.

And, while I'm editing, don't get me started on bulbs that emit 300-400 lumens being advertised as 60 watt replacements. 850 is a 60 watt replacement, though I'll concede 800 which is what several manufacturer use as it is the energy start standard. The advertising is bull in many cases, and my view is rather jaundiced at this point.

That's how capitalism works.

 

[mhaze]

Repeating from the middle of my previous rambling post, one interesting data set.

The stuff I put in in 2006 has ran nonstop since then, 5 years x 24 x 360 = 43,200 hours. One failure of about 8 or 10 total

 

[roger]

I could go take my point and shoot thermometer and provide data as to temperature in the interior of a surface mount ceiling can with a 40 and a 65 watt LED, but I know without doing this that the temps in there don't get up to levels that would seriously affect electronics. But the argument against LED in a ceiling can based on heat accumulation seems to be an argument in the direction of "the heat sink doesn't work or isn't engineered properly", which is a bit ridiculous.

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.

Please make sure you tell the distributor if you will be using the light in a recessed can as the light needs to be de-rated to 10 Watts for a 5" or 6" recessed can or to 8 Watts for a 4" recessed can. If you do not use a de-rated light in a recessed can the warranty will not no longer be valid.

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?

 

[roger]

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?