Nichia Develops 60 Lumen Per Watt White LED

[jtr1962]

jtr,

Do you have comments on this article? Is it possible that we're on verge of a massive cost reduction in LED production costs?

I wouldn't say we're on the verge of making LEDs much cheaper, but this is a step in the right direction. Let's not forget that the die isn't the most expensive part of an LED. The package is what costs the most. Power LED packages are coming down in price but still high relative to the die. Then again, if the dies were on standard silicon then perhaps this would make the packages less expensive as well.

It would be very nice if LED lighting became cost effective for general use (i.e. cheaper than all the alternatives).

Even without this development, we're already there for small lighting uses. LED Christmas light prices were pretty close to incandescent this year. I suspect in 2009 or 2010 they will reach parity, and incandescent Christmas lights will disappear. For general lighting however we still have a way to go. Even if the LEDs cost nothing, it is necessary to convert 120 VAC into something the LEDs can use, and also to heat sink them so they reach their rated lifetimes of 50,000 to 100,000 hours. Both of these things right now drive up the prices of LED lighting well past the point of incandescent or even CFLs. Granted, when looked at from a total life cycle point of view a $25 LED bulb easily pays for itself, but many consumers only buy based on initial purchase price. Still, I feel once we get the price of LED bulbs under $20, perhaps even under $10, and educate consumers about the money they'll save long term, then LED can compete with the alternatives. However, we're going to have to make a quality lamp for this price-one which is dimmable, available is the common color temperatures, lasts as long as advertised, and has little variation in tint between bulbs. Right now I see too many LED screw-in lamps which are cheap junk. I hope this doesn't give a bad name to LED. Note however that LED screw-in bulbs are only an interim solution. Longer term, it'll make more sense to sell purpose-built LED fixtures which are designed to last a lifetime without replacing lamps. For many reasons, screw-in lamps are a poor form factor which requires too many design compromises in terms of space, heat dissipation, etc.

 

[sechs]

That's a little optimistic. Kind of like saying VOIP is an interim solution until cell phones can really replace analog land-line phones.

You can't truly expect people to replace most, if not all, of the lighting in their house just so that they can use LEDs, no matter how much money they might save. It doesn't matter how suboptimal the edison base is, the masses will never switch unless the screw-in replacement LED lamp is perfected.

How long have CFLs been around, and only now are they making real penetration? And what percentage of those are *not* screw-in replacements?

 

[LunarMist]

I agree. LEDs will have to be better than CF lamps, and that means soft light covering 360x270°, a wider range of color temp options than typical white LEDs, models covering the 500-2500 lumen range, and equal or better efficiency and cost, life, etc. I would avoid buying a lamp with fixed lights of any kind. Readily available replacement bulbs are a must. (I think Edison figured that out.) I've already seen how quickly the LED nightlights and crappy Christmas lights dim, flicker or fail over a relatively short period. I have some fluorescent lights that have been on 24/7 for ~3 years, and some others that lasted 5-7 years in typical use. LEDs are not there yet.

 

[jtr1962]

I've already seen how quickly the LED nightlights and crappy Christmas lights dim, flicker or fail over a relatively short period.

Those all use 5mm indicator-type LEDs which at best dim to 50% brightness after a few thousand hours (really bad ones can dim in a few weeks). Power LEDs like Crees or Luxeons use silicon encapsulant and really do last 50,000 hours or more. I have a Luxeon which has been running since January 2004. It hasn't dimmed appreciably.

If an LED fixture is properly designed then there doesn't need to be any provision for replacing the emitters. And while I agree most people won't immediately replace all their fixtures to use LED (hence you will need LED screw-in bulbs even if they're suboptimal), remember that fixtures do have a finite lifetime. A lot of people do in fact change fixtures when they change decor. That would be when they might replace a screw-base fixture with a purpose-designed LED fixture.

Interestingly, those designing LED streetlights aren't making any provisions to replace the lamps. The idea is to design the light source with a lifetime of perhaps 1 million hours by underdriving the LEDs. By doing this the LEDs will last the lifespan of the streetlight fixture (about 200 years).

 

[LunarMist]

Unfortunately most Chinese consumer products use the cheapest, crappiest circuits and components possible. I don't know the budget for street lights, but 200 years is rather unrealistic. In many areas the poles and infrastructure would not last that long, and I doubt that the circuits would hold up. Besides, they will most likely be replaced with newer technology in less than 50 years. Around here bad drivers are the death of many a light standard.

 

[jtr1962]

I know for a fact some parts of NYC have streetlights over 100 years old. And those cobra-style lights which are just about everywhere date from the late 1940s/early 1950s (except the ones destroyed by drivers). Most still have decades of life left in them. 200 years is a long time but not totally unrealistic or unprecendented. Most modern skyscrapers are designed with a 200-300 year lifespan in mind, for example. Regarding the electronics, if everything is sealed and overdesigned you can in fact make it last a very long time. I still have a digital alarm clock dating from the late 1970s which works fine. Naturally, overdesigning means higher initial costs but the tradeoff is little or no ongoing maintenance. With state budgets being stretched thin, the idea is to get infrastructure projects as cost effective as possible long term even if it costs more up front. Remember that states float bonds to pay for projects. They might pay for the new, more expensive streetlights over a period of 30 years (versus perhaps 20 for a less expensive version). Same amount per year out of the budget. The difference is with the less expensive version once the initial bond is paid off, maybe you'll have to float another bond to pay to replace them. The more expensive version will be just fine long after the bonds are paid.

Besides, they will most likely be replaced with newer technology in less than 50 years.

This is why municipalities haven't yet jumped on the LED streetlight bandwagon. They don't want to buy streetlights when in perhaps a year or two you'll have ones 50% more efficient. Instead they're waiting a while until LED technology starts to plateau, and also using up the full life-cycle of existing equipment. This will probably occur in a few years when we reach 150 to 200 lm/W. After that, it can't get much better, at least not with blue plus phosphor LEDs. Also, they're now learning how to get the longest life out of LEDs from the existing installations of LED traffic lights. This knowledge will be readily appied to LED streetlight design.

 

[LunarMist]

Meanwhile, where is my TK40 already?

 

[ddrueding]

I'm projecting a full-house-gutting remodel in about 2 years. I wonder if LED will be ready for a full-custom install by then? Should I plan on wiring the house with 24v circuits for lighting?

 

[udaman]

I'm projecting a full-house-gutting remodel in about 2 years. I wonder if LED will be ready for a full-custom install by then? Should I plan on wiring the house with 24v circuits for lighting?

2yrs ago, the mortgage foreclosure crisis had not gone nuclear...2yrs is a long time. If you have enough money, LED is ready now, all depends on what your usual unspecified qualification issues are . Wait until 3mo before you decide to do it, to then actually waste anytime thinking about it. Ergo, I'm thinking of upgrading to a new Mac laptop & SSD drive, do you think laptops and SSD's will be good enough by then?

Aside, hope LM didn't by the Fenix from that Chinese Christian spammer DB '4sevens'...or was that a laser copier/printer LM was referring to .

 

[jtr1962]

I'm projecting a full-house-gutting remodel in about 2 years. I wonder if LED will be ready for a full-custom install by then? Should I plan on wiring the house with 24v circuits for lighting?

It's a good idea to put in low voltage DC circuits regardless. If you ever go with solar panels then you can use the panel output directly for things like lighting, without using an inverter to convert to 120VAC. I think the standard for compatibility with solar panels would be 48VDC but I'll have to double check.

As udaman said, you can do a house with LED now if price is no object. In two years it'll be even better and cheaper. High CRI LEDs are the next thing.

 

[LiamC]

I noticed that Canberra was replacing incandescent traffic lights (par 56/64?) with LED based two or three years ago. I see a significant number of these with a significant amount of blown emitters. Green seems to be worse than red/orange

 

[ddrueding]

I do need high-CRI, I've become quite spoiled by 6500k/95+ CRI florescent bulbs and lots of light. That makes LEDs even more expensive.

 

[LunarMist]

Yuck, I'd not want to live under 6500K lights. To me 4000K is for living and 5000K is for (D50) working.

 

[ddrueding]

6500K is for working, I find that I stay awake longer and easier with them. 5000K for at home, and 4000K for outdoor lighting at night.

 

[RWIndiana]

I've recently developed an interest in LED lighting myself. I'm looking for some decently bright C7s (ie 15-25 watt equivalent), not just the ordinary Christmas light type (those are very dim). Anyone know of some like that?

 

[ddrueding]

74 Lumen / Watt at respectable power levels (7400 Lumens, 100 Watts)

Gotta get me some of these...

 

[P5-133XL]

Are you sure it isn't the eye glasses you covet?

 

[ddrueding]

I think I'm good, thanks

 

[ddrueding]

Panasonic 60W(equivalent) LED bulb. 19 year lifespan for $40?

 

[Stereodude]

Assuming it lasts 40 years.

We all know how CFL bulbs don't last anywhere near as long as they claim, why would we fall for it this time?

 

[Will Rickards]

my cfls have lasted 5+ years. A couple have failed early (less than 1 year) but those were failures in the base of the bulb that looked like arcing occurred.

 

[mubs]

All my CFLs I have broken. It seems they use the thinnest of glass to make the tubes. Putting them in and taking them out of bayonet mount lamp holders breaks them really easily. It is not always possible to hold the bulb by the base when trying to put them in certain light fixtures. Good old incandescents were strong enough to fix / remove them multiple times while holding the glass portion.

 

[sechs]

Assuming it lasts 40 years.

We all know how CFL bulbs don't last anywhere near as long as they claim, why would we fall for it this time?

It's 19 years, and, if you use the bulb the way that they describe, you should get reasonably close.

I, for one, don't use lights as most manufacturers rate their lamps. Yet, even my cheap CFLs last a year in the refrigerator.

 

[Stereodude]

Either way, I don't think the electronics driving the LED will last 19 years.

 

[LunarMist]

My CFLs are not lasting very long either. The problem is that they are designed to be on for hours at a time and the lifetime is definitely reduced with numerous short cycles. (The warmup period sucks, too.) I have one 9W on 24/7 in the kitchen hood and it lasted about 2.5 years, which is not bad. Others are lasting about 1-2 years with 1-3 hours of use per day.

I have not seen breakage as Mubs experienced.

 

[sechs]

Either way, I don't think the electronics driving the LED will last 19 years.

How long do you think they will last, then?

 

[sechs]

Others are lasting about 1-2 years with 1-3 hours of use per day.

If you're buying cheapies, I think that 2 to 3 years is a reasonable life under real use conditions.

I think that there's something to be said for buying higher quality CFLs. They really do last longer when used in my fridge.

 

[P5-133XL]

My CFL's have all lasted more than 5Y and on all but bathroom they have been going strong for over 8Y. For some reason the ones above the mirror don't survive well.

 

[LunarMist]

If you're buying cheapies, I think that 2 to 3 years is a reasonable life under real use conditions.

I think that there's something to be said for buying higher quality CFLs. They really do last longer when used in my fridge.

My refrigerator does not have any exterior lights.

 

[Stereodude]

How long do you think they will last, then?

That depends on the driving circuit of the LED and how cheap the components are in it. The ballast in CFL's is what fails because they're super cheap. If a LED bulb uses something equivalent they won't last their rated lifetime either.

 

[Fushigi]

My CFL's have all lasted more than 5Y and on all but bathroom they have been going strong for over 8Y. For some reason the ones above the mirror don't survive well.

CFLs in a bathroom are most likely going to suffer from lots of short duty cycles. Also, for full baths, does the higher humidity impact their life?

 

[ddrueding]

My refrigerator does not have any exterior lights.

I don't know of any that do. I suspect he was talking about the internal light. That and the kitchen hood are both fairly harsh environments (humidity, temp extremes, etc).

 

[Fushigi]

The vending machine at our downtown office lights the snacks via LED. Not too innovative in and of itself, but the lights are motion activated and don't come on until you're about 6 feet from the machine.

 

[Stereodude]

I see a lot of white LED lights in the glass door freezer units at the grocery store as of late.

 

[LunarMist]

I don't know of any that do. I suspect he was talking about the internal light. That and the kitchen hood are both fairly harsh environments (humidity, temp extremes, etc).

I've replaced maybe 5 refrigerator lights in my life. Why bother with a CFL replacement in such limited use?

 

[sechs]

My refrigerator does not have any exterior lights.

I guess that if they were on the exterior, then they would be in it, would they?

 

[sechs]

That depends on the driving circuit of the LED and how cheap the components are in it.

This is to say that you're talking out of alternative orifices.

 

[sechs]

That and the kitchen hood are both fairly harsh environments (humidity, temp extremes, etc).

My hood CFLs have only failed due to user error and old age.

But they do get kind of gunky, and you have to clean them regularly.

 

[sechs]

I've replaced maybe 5 refrigerator lights in my life. Why bother with a CFL replacement in such limited use?

They're brighter at lower wattage. They also come in daylight, rather than just sickening yellow.

If LEDs weren't so stinkin' directional, I'd use them instead, as they take the cold better and produce even less heat.

 

[Howell]

I have one 9W on 24/7 in the kitchen hood and it lasted about 2.5 years, which is not bad. Others are lasting about 1-2 years with 1-3 hours of use per day.

I would be hesitant to place something that has mercury in it so close to my food and heat.