News from the largest LED fair: Cree, Nichia, Philips, Seoul new LEDs

[ledstein]

Light+Building fair in Frankfurt Germany is the worlds biggest lighting fair, held every two years. I paid a visit and thought to share the important news from the best known LED producers:

Cree:

- New and improved MT-G, called MT-G version II. The details are few but it seems it will have 54% more lumens at 25W, 2586lm instead of 1670lm
- XP-G is back with a twist: smaller size, better price and is now called the XB-D. We will have to wait until summer to see what "better price" means.

Philips - Lumileds:

- Remember K2? Now we have K 4, 8, 12, 14, 16 and 24. The number means the number of LEDs in a module. The K 12 is similar in performance to the MT-G II.


Nichia:

- No new LEDs, only improvements on the current versions: 183, 219 and 157.

Seoul Semiconductor:

- Z series, Z5 - similar to Cree XP-E, Z7 - similar to Cree MC-E
- ZC their first COB LED


At www.facebook.com/LedRise you can see some pictures from the fair, a comparison between XP-G and the new XB-D, MT-G Gen II vs K 12 UP and what Nichia improved to their LEDs.

[znomit]

Cree:

- New and improved MT-G, called MT-G version II. The details are few but it seems it will have 54% more lumens at 25W, 2586lm instead of 1670lm

Want!
http://www.cree.com/led-components-a...nal/xlamp-mtg2

[ledstein]

Yes MT-G II looks very interesting. One important development which can be seen also in the little information that Cree shows about this LED: gone are the days in which the performance was measured in the lab at 25C. Both Cree and Philips now put numbers closer to the real life performance in their new datasheets: at 85C.

[ledstein]

Cree demonstration modules: Zhaga.

- 3690 lm from 32 XB-D LEDs packed in a 45x45mm cluster with 37W, 99lm/w. Now thats a PUNCH!

http://www.cree.com/led-components-and-modules/zhaga
https://www.facebook.com/photo.php?f...type=1&theater

[RoGuE_StreaK]

- XP-G is back with a twist: smaller size, better price and is now called the XB-D.

Apples and oranges I would have thought? XB-D is only rated to 1Amp, and highest spec is 287Lumens at that current, whereas the XP-G is rated to 1.5Amp with highest spec giving 463Lumens?
If anything, I would have thought the XT-E White was the "new XP-G". XB-D seems to be aimed more at arrays rather than individual sources?

[ledstein]

Apples and oranges I would have thought? XB-D is only rated to 1Amp, and highest spec is 287Lumens at that current, whereas the XP-G is rated to 1.5Amp with highest spec giving 463Lumens?
If anything, I would have thought the XT-E White was the "new XP-G". XB-D seems to be aimed more at arrays rather than individual sources?

Well that is what the Cree sales guy at the stand said. Also the XT-E is the successor of the XP-E as we can read online too:
"The Cree XT-E LED matches the widely-used XP-E LED in footprint and price while offering double the efficacy based on the company's newest silicon-carbide manufacturing platform."

http://www.ledsmagazine.com/news/9/2/7

A successor of a LED does have to be a similar LED. If i remember correctly, the XP-G replaced the XR-E while being at the same time smaller and with better performance. Now the next step is: even smaller and with better price...

[ledstein]

Zhaga (http://www.zhagastandard.org/) is also one of the big news from Light+Building. Zhaga is a global organization that has a very noble purpose for us consumers: to enable interchangeability of LED light sources made by different manufacturers. Most of the big names in the industry are already part of the 190+ membership.

But will work? Will we be able to buy a LED product (torch, lamp, fixture) and later upgrade the LED or replace it if it burns up?

In my opinion Zhaga will not work because of the following reasons:

1) Its is against the rules of profit. Why give the consumer the option to pay 500 USD (for example) for a trendy LED fixture and then let him upgrade its LEDs for a much lower cost when you can make him buy the hole fixture again? Its like Apple making the Ipad easily upgradable from the chip point of view and not making us pay hundreds of dollars over and over for a new product. Even in the industry that invented the trend to upgrade, the computer industry, the most sold product today is not upgradable: the laptop. You want better graphics or a faster CPU? You buy a new laptop!

2) The need to make interchangeable light sources and to set standards for shape came from the very short lifespan of the traditional light sources. If you use a incandescent that has only 1000 hours of life you really need to make it easily replaceable. LEDs last much longer, too long to make this need really important.

3) Not even standards last long. Look again at the computer industry: my desktop is 5 years old and, in practice, i mostly cannot upgrade it anymore. The new CPUs or memory do not match my motherboard. With LEDs will be the same, the standards of today will be obsolete in a few years...

4) Everybody can change a bulb but LEDs, even if interchangeable modules exist they will be fixed deep inside the light fixture so most will of us will not be able to change them anyway...

5) There will be way too many standards and it will be like they do not exist at all.

[HumanLumen]

I think Rogue Streak is right WRT the XB-D. The XB-D already exists and is availble now and is the DA1000 1 x 1mm die - the XP-G is 1.4mm x 1.4mm. Maybe you misheard him say XB-G (based on SC3 technology - that would be cool/sweet!!)?
HL

[idleprocess]

Zhaga (http://www.zhagastandard.org/) is also one of the big news from Light+Building. Zhaga is a global organization that has a very noble purpose for us consumers: to enable interchangeability of LED light sources made by different manufacturers. Most of the big names in the industry are already part of the 190+ membership.

But will work? Will we be able to buy a LED product (torch, lamp, fixture) and later upgrade the LED or replace it if it burns up?
[...]
5) There will be way too many standards and it will be like they do not exist at all.

I think this will be the biggest problem with Zhaga. Looking at their preliminary standard, it looks like every OEM will soon have a nice set of Zhaga-approved standards to live within that are conveniently incompatible with every other OEM.

Their apparent difficulty settling on fairly simple terms such as "light engine" is also troubling - on CPF, a "light engine" features LED(s), control circuitry, and a heat spreader for interfacing with thermal management. But for Zhaga, it could be almost any logical combination of LED, heat spreader, thermal management, control/driver circuitry, and optics - anywhere from being analogous to a light bulb in an incandescent fixture to a nearly-complete fixture that just snaps into the housing (such as the next revision to the Bridgelux Helion module).

I personally think that it's a waste because of another point:

2) The need to make interchangeable light sources and to set standards for shape came from the very short lifespan of the traditional light sources. If you use a incandescent that has only 1000 hours of life you really need to make it easily replaceable. LEDs last much longer, too long to make this need really important.

I think that the vast majority of general-purpose interchangeable LED lighting will be incandescent replacements. The purpose-built stuff will be treated like light switches, electrical outlets, and plumbing fixtures - lasting so long as to render quick-change an unnecessary expense. While there are some environments like retail where it will likely catch hold due to the need to follow trends or simply change up displays seasonally, places like offices and homes likely won't need so much flexibility.

Businesses will likely buy "monolithic" LED lighting once the manufacturers have some long-term test data and the cost of acquisition gets closer to florescent. They'll like the "set it and forget it" of LED that's also nearly maintenance-free, as opposed to floro that requires routine tube replacements and occasional ballast swaps, driving up its operational expense. I suspect that LED modules are already much more cost-competitive on a true TCO basis, but since the maintenance costs aren't part of the capital expense, they can be largely ignored when building new facilities - especially when leased out (leaving routine maintenance to the tenant) or sold to the final occupant.

[blasterman]

I think that the vast majority of general-purpose interchangeable LED lighting will be incandescent replacements.

Good point, and similiar to the one I've been making in teh fixed lighting forum. Comercial customers are going to make much longer term decisions based on hard numbers and a 'set it and forget' mentality. Because something better comes along doesn't mean they'll swap out the esisting fixture unless it proves itself on a spread sheet it can save money.

Residential / Consumer on the other hand has different expectations given that consumers are likely to replace functional products often for emotional reasons, even if it's the perception of saving money -vs- reality. The LED market is quick to play to those emotions. "Well, you bought a 50,000 hour LED bulbs last year? Well, you need to replace it because our 50,000k hour LED bulb is better".

The commercial sector is being very fickle about upgrading to LED based fixtures, especially when it comes to interiour lighting. I'm seeing a rapid increase in recessed ligting in homes converting to LED because the fixtures (LR6 / CR6 for example) are getting cheap and the light quality rivals halogen. I'm not seeing this happen in the commercial sector with any momentum because there need to be a more justifed reason to tear out T8 fixtures.

The writing on the wall is Western Lighting companies have to come up with something to impress the commercial sector because the Chinese manufacturers are following closely behind with less efficient but radically cheaper fixtures.

[idleprocess]

The commercial sector is being very fickle about upgrading to LED based fixtures, especially when it comes to interiour lighting. I'm seeing a rapid increase in recessed ligting in homes converting to LED because the fixtures (LR6 / CR6 for example) are getting cheap and the light quality rivals halogen. I'm not seeing this happen in the commercial sector with any momentum because there need to be a more justifed reason to tear out T8 fixtures.

I suspect this is a result of the capital budget vs maintenance cost dilemma that corporations have faced since the beginning of the free market. A high quality / more efficient / more durable / etc good costs appreciably more upfront, but produces better quality output / requires less input to operate / lasts longer / etc than a lower-quality good. This problem is exacerbated when organizational incentives encourage small-horizon planning and those responsible for the capital budget are rewarded for conserving it rather than on the basis of TCO. In other words, short-sighted use of capital often robs Peter to pay Paul.

I don't see any ground swell movement towards LED in the next few years for a number of reasons:

• Given the sheer number of T12 fixtures I've seen used in office and academic environments in recessed and suspended fixtures for the ubiquitous suspended-ceiling commercial office building / university classroom, it's no surprise that the newer T8 & T5 fixtures are still going strong
• Businesses have internalized the maintenance costs as normal
• Crony capitalism likely has more than a few brother-in-laws performing low-profile work like building maintenance contracts
• Wages are usually a company's largest expense; the slightest change in the office environment that someone whines about may well result in a real loss of productivity that erases the savings from an efficiency enhancement like new lighting
• The tough market is encouraging businesses to extend the service life of all equipment as much as possible
• In my experience, the folks that specify equipment for office buildings tend to be conservative, unimaginative, and far too busy to experiment with anything new; they may also face severe disincentive for rocking the boat or committing errors; Floro is mature and almost universally well-known - can't say the same for LED
• Electricity is so cheap and lighting such a small percent of a commercial building's energy cost that it's not a priority

Change a few variables and give the LED lighting market some time to mature and perhaps things will change. I suspect things to start changing once the manufacturers stabilize their offerings, start to show some hard numbers on long-term testing, and some of the name brands start offering product. I imagine that it will be a steadily-increasing trickle as LED eats into the new-building/renovation/retrofit/replacement market as old floro fixtures start to expire. Plenum suspended ceilings are an OK place for LED fixtures anyway...

[zzonbi]

Cree MT-G2 has better specs than CXA2011, but doesn't seem available. It's using the thick pyramid die tech of the xt.
There's also a Seoul 3535 monodie led in the 3.5mm class - xp, nichia119/219 - at interesting prices at led1.de

[ledstein]

I personally think that it's a waste because of another point:

I think that the vast majority of general-purpose interchangeable LED lighting will be incandescent replacements. The purpose-built stuff will be treated like light switches, electrical outlets, and plumbing fixtures - lasting so long as to render quick-change an unnecessary expense. While there are some environments like retail where it will likely catch hold due to the need to follow trends or simply change up displays seasonally, places like offices and homes likely won't need so much flexibility.

Businesses will likely buy "monolithic" LED lighting once the manufacturers have some long-term test data and the cost of acquisition gets closer to florescent. They'll like the "set it and forget it" of LED that's also nearly maintenance-free, as opposed to floro that requires routine tube replacements and occasional ballast swaps, driving up its operational expense. I suspect that LED modules are already much more cost-competitive on a true TCO basis, but since the maintenance costs aren't part of the capital expense, they can be largely ignored when building new facilities - especially when leased out (leaving routine maintenance to the tenant) or sold to the final occupant.

Incandescent replacements are only a transition phase for LEDs as they are trying to reach "critical mass" needed ,in term of volumes, to be price competitive and at the same time make us more knowledgeable about their use as lighting products. From the design point of view, and we usually buy a fixture for its design or label, its nonsense to keep revolving indefinitely around the shape of a, soon to defunct, light source.

What Zhaga is trying to achieve is very good for us consumers but their approach is wrong. We really do need to be able to prolong the life of a LED fixture beyond the life of the LED.

- For residential lighting, a lighting fixture can have a useful life of decades or even centuries as i embodies an era in our history. There are plenty of homes that furbished in the style of early 1900s or 1960s and so on. Maybe in fifty or a hundred years or something people will want to have a home in the early 21st century style and that implies lighting fixtures too. But wait, you wont be able to use these fixtures as their light sources do not function anymore and they cannot be replaced...
- For commercial and business lighting a lifetime of 25.000 hours can mean only 2-3 years of operation. It is not a sound business practice to replace hole fixtures because the LEDs do not function anymore or are too dim, especially considering the cost of the LEDs can be only a fraction of the cost of the fixture.

In my opinion Zhaga should do two things:

1) Standards of LED shape and current. Even if they will be 10, 20 standard shapes there is a pretty good chance they will last a long time. When the LED burns up you take the fixture to a service center of lighting products (yes, they need to exist first). This way of doing will work for residential market or for business that do not use plain lighting fixtures but those that are more expensive and have a better design.
2) A handful of standards for light engines for markets where huge volumes of the same or similar fixtures are used. Office lighting and street lighting are the main candidates. In these markets functionality is much important than design.

[ledstein]

I don't see any ground swell movement towards LED in the next few years for a number of reasons:

• Given the sheer number of T12 fixtures I've seen used in office and academic environments in recessed and suspended fixtures for the ubiquitous suspended-ceiling commercial office building / university classroom, it's no surprise that the newer T8 & T5 fixtures are still going strong
• Electricity is so cheap and lighting such a small percent of a commercial building's energy cost that it's not a priority

LEDs where trying to take on the neon in T8 or T5 fixtures in the wrong way. The LED tube has been around for years but did not really gain any ground as you cannot really make a tube filled with electronics and metals as cheap as a glass one filled with gasses. Now the approach has changed as i have seen at the fair. T6 or T8 tubes are to be replaced with modules as they ones here (Philips Fortimo):

http://www.lighting.philips.co.uk/su...es-systems.wpd

or even strips:
http://www.ledrise.com/led-lighting-...wer-led-strip/

Both much cheaper to build than the LED tube. And considering the huge number of Chinese "visitors" with huge cameras that snapped pictures at every product at the lighting fair we will see tons of cheap Chinese copies, very soon.

Lighting represents about 15-30% of the energy cost and 5-15% of the heating cost of a commercial building.

[idleprocess]

What Zhaga is trying to achieve is very good for us consumers but their approach is wrong. We really do need to be able to prolong the life of a LED fixture beyond the life of the LED.

I don't think LED's themselves will ever be light bulb, HID module, nor floro-cheap for quite some time - the cost of the LED's themselves will continue to be a sizable percentage of the fixture costs.

- For residential lighting, a lighting fixture can have a useful life of decades or even centuries as i embodies an era in our history. There are plenty of homes that furbished in the style of early 1900s or 1960s and so on. Maybe in fifty or a hundred years or something people will want to have a home in the early 21st century style and that implies lighting fixtures too. But wait, you wont be able to use these fixtures as their light sources do not function anymore and they cannot be replaced...

That's like worrying about the plight of film photographers now that digital has taken over. Their hobby was more affordable before digital because everyone else was (grudgingly) subsidizing them. Once the market went digital only the diehards stuck with it, and with volumes down they're paying more. Eventually the capital cost of new digital equipment will be more appealing than the spiraling operational costs of buying and processing film.

For every centerpeice lighting fixture that's worth something aesthetically, there are a dozen cheap builder-grade fixtures with approximately zero aesthetic or sentimental value. Think fixtures in closest, hallways, utility rooms, garages, spare bedrooms, secondary bathrooms, etc ... their sole function is to provide light, reliably.

In fifty or 100 years if you want to decorate your house in the style of the late 20th century, you will either need to visit antique stores for the real deal or turn to specialty manufacturers for something appropriately retro... and expect to cough up a small fortune for either. I have no real interest in ensuring that such fixtures are affordable to obtain and operate in the future. CFL's and LED bulbs will be around for numerous decades to serve the existing fixture stock; much longer than that and attrition will take over.

- For commercial and business lighting a lifetime of 25.000 hours can mean only 2-3 years of operation. It is not a sound business practice to replace hole fixtures because the LEDs do not function anymore or are too dim, especially considering the cost of the LEDs can be only a fraction of the cost of the fixture.

All the commercial LED fixture manufacturers I'm seeing are promising appreciably more than 25k hours. Only thing I see that claim on are LED bulbs, which live in decidedly poor environments. Given that a number of 24-hour operations around me have switched to LED lighting, I will be watching their maintenance patterns closely.

Yes, the trend in industry towards treating everything as a sealed assembly is maddening. The reduction in manufacturing cost is paid by the unfortunate consumer/operator that has to deal with a component failure that cannot be satisfactorily diagnosed nor repaired short of specialized, warranty-voiding labor and unsupported replacement parts.

With the interest on high temperatures in binning for the newest LED models from the major manufacturers, increasing environmental ruggedness, and even new die fabrication techniques that seem to be addressing some of the shortcomings of present LED tech, I expect manufacturers to start hitting 100k+ hours easily in a few years - especially with active thermal management. That would be whole-fixture replacement at more than a decade interval.

For commercial/industrial applications, some ability to replace wear components (LED's, driving circuitry, active thermal management) will be a selling point... but I don't expect said interchangability to be as ubiquitous as the A19 bulb nor the 48" T8 floro. I suspect it will be more like automobile parts - the OEM will support their wares so long as it's profitable with lower-overhead third-party aftermarket manufacturers at the fringes supporting smaller-volume parts where they can make a profit. Each manufacturer will have a variety of "platforms" with variable degrees of interchangeability between models within a "platform".

Once the fixtures can comfortably reach 100k hours - largely achieved by superior efficiency and declining unit costs that allow much gentler driving for similar "lumens per dollar" - I'm not sure that the residential sector will need the modularity. I see 3-4 hours a day listed as a common metric for light bulbs when determining lifespan. If we extend that to 8 hours a day, a 100k hour operating lifespan would realize about 34 years of operation - longer than is typically realized for numerous other hard-wired / permanently-installed household equipment such as climate control and water heaters.

LEDs where trying to take on the neon in T8 or T5 fixtures in the wrong way. The LED tube has been around for years but did not really gain any ground as you cannot really make a tube filled with electronics and metals as cheap as a glass one filled with gasses. Now the approach has changed as i have seen at the fair. T6 or T8 tubes are to be replaced with modules as they ones here (Philips Fortimo):

http://www.lighting.philips.co.uk/su...es-systems.wpd

or even strips:
http://www.ledrise.com/led-lighting-...wer-led-strip/

LED strips are nothing new and seem to largely be the domain of sign companies replacing neon or doing custom outdoor lighting for retail customers. Or hobbyists such as myself that can't be bothered with too much fabrication for less aesthetically-demanding applications.

The Philips offering seems to be a semi-transitional offering available to other lighting manufacturers so as to minimize their R&D costs on fixture design since their strip fits within the general footprint of standardized linear floro systems. Namely, "chunk this strip/heat spreader/driver into your existing footprint then figure out optics and thermal management." Suspect that would make fixture manufacturers somewhat uneasy, much like Intel's foray into motherboards more than a decade ago was initially upsetting to the PC OEM's at the time.

[Lux-RC]

I don't think that XB-D is a direct replacement of XP-G
At least for the flashlight domain.
XB-D (like XT-E and XP-E HEW) has probem with color uniformity which makes it almost useless for projection optics (like TIR collimators and reflectors).
Also, it's limitted at 1A maximum current.

Cheaper doesn't mean better, it least in this case.

From my stand point it looks more like a step back.

[EricB]

I don't think LED's themselves will ever be light bulb, HID module, nor floro-cheap for quite some time - the cost of the LED's themselves will continue to be a sizable percentage of the fixture costs.

Aren't LED's cheaper to make, since they're solid state? I thought it was just a matter of the cost of the newer GaN coming down, and then LED's would be the all-around most cost-effective solution.

[BrightLED17]

Where can I buy the Seoul LEDs?

[RoGuE_StreaK]

LedRise seem to have them, I'm sure any number of other places do as well.

[ledstein]

I don't think LED's themselves will ever be light bulb, HID module, nor floro-cheap for quite some time - the cost of the LED's themselves will continue to be a sizable percentage of the fixture costs.


Once the fixtures can comfortably reach 100k hours - largely achieved by superior efficiency and declining unit costs that allow much gentler driving for similar "lumens per dollar" - I'm not sure that the residential sector will need the modularity. I see 3-4 hours a day listed as a common metric for light bulbs when determining lifespan. If we extend that to 8 hours a day, a 100k hour operating lifespan would realize about 34 years of operation - longer than is typically realized for numerous other hard-wired / permanently-installed household equipment such as climate control and water heaters.


LED strips are nothing new and seem to largely be the domain of sign companies replacing neon or doing custom outdoor lighting for retail customers. Or hobbyists such as myself that can't be bothered with too much fabrication for less aesthetically-demanding applications.

The Philips offering seems to be a semi-transitional offering available to other lighting manufacturers so as to minimize their R&D costs on fixture design since their strip fits within the general footprint of standardized linear floro systems. Namely, "chunk this strip/heat spreader/driver into your existing footprint then figure out optics and thermal management." Suspect that would make fixture manufacturers somewhat uneasy, much like Intel's foray into motherboards more than a decade ago was initially upsetting to the PC OEM's at the time.

50k hours, 100k hours lifetime was from the start the main sales pitch of LEDs and was diligently used for any fixture of any kind that had LEDs inside. For a while such claims where untrue for 99% of the products, then the serious producers started to back down from them and today only Chinese no-names still use them every day. Most fixtures from Samsung, Toshiba, Osram, Philips and many more i have seen have a claimed lifetime of 15k,25k with some exceptions in the 40k area. In all cases we need to be consider the fine print (many times not even mentioned): such lifetime is possible only if certain conditions are met, most importantly, those concerning ventilation and ambient temperature. Too bad in LED fixtures they did not put some sort of material that changes color, or smt, when exposed to a certain temperature for a prolonged interval of time. Then we could know if the fixture malfunctioned because of the heat or a flaw in the system design.

That is why some light needs to be made in this aspect of interchangeability, LED fixtures in most cases do not last as much as claimed. Do not get me wrong, i am fan of LEDs, they are the future but the road to there is quite bumpy.


About strips, they have a much important role as you think. They are simple products, easy to service and produce and with increasingly better lumen/dollar ratio. A large slice of the lighting market is:

- linear lighting for residential: http://bit.ly/J0g49X
- linear lighting in office with the common fixture shapes (60x60, 120x30 and so on): http://bit.ly/J0g49X

For both, LED strips (if they have high flux or high efficacy) are perfect. In residential, the light source is hidden so no need for a complex fixture, for office functionality matters a lot more so again no use to build a case around it (as they do it with the LED tube). The Philips strips in question do not look like some parts of a fixture build by a 3rd party, they are just meant as light source to replace another light source, the neons.

Here is some idea where the prices are today for strips:

http://www.ledrise.com/led-lighting-...wer-led-strip/
http://www.ledrise.com/led-lighting-...ble-led-strip/

[idleprocess]

50k hours, 100k hours lifetime was from the start the main sales pitch of LEDs and was diligently used for any fixture of any kind that had LEDs inside. For a while such claims where untrue for 99% of the products, then the serious producers started to back down from them and today only Chinese no-names still use them every day. Most fixtures from Samsung, Toshiba, Osram, Philips and many more i have seen have a claimed lifetime of 15k,25k with some exceptions in the 40k area. In all cases we need to be consider the fine print (many times not even mentioned): such lifetime is possible only if certain conditions are met, most importantly, those concerning ventilation and ambient temperature. Too bad in LED fixtures they did not put some sort of material that changes color, or smt, when exposed to a certain temperature for a prolonged interval of time. Then we could know if the fixture malfunctioned because of the heat or a flaw in the system design.

No, they're not there yet, but I believe that they will get there. With the new binning specs at 85C and some of the new substrate tech being researched, I believe it won't be long until we see useful lifespans under real-world conditions at real-world drive currents at 100k hours.

Your Philips link is claiming 50k hours without no heatsinking beyond the strips themselves.

That is why some light needs to be made in this aspect of interchangeability, LED fixtures in most cases do not last as much as claimed. Do not get me wrong, i am fan of LEDs, they are the future but the road to there is quite bumpy.

I'm sure we'll see it, but I seriously doubt it will lead to lower prices nor near-universal standards. If industry has learned anything in the last several decades it's that proprietary standards let them sell at higher margins (even though reasonably-priced universal standards lead to greater volumes and potentially greater overall profits). The financial talking head types love high margins - even if unsustainable - thus that's the focus in industry these days.

Even the cell phone manufacturers - notorious about making accessories proprietary - haven't really caved on using USB as a universal standard... there are still a variety of connectors and some will charge at an infuriatingly low rate when using non-OEM sources (well below what the source can supply).

About strips, they have a much important role as you think. They are simple products, easy to service and produce and with increasingly better lumen/dollar ratio.

Jury's still out on cheap LED strips. I gather that they have high failure rates. I installed some recently; if they go out no big deal for me since I'm a hobbyist ... most consumers will not be so understanding.

They are interesting in the sense that with greater LED efficiencies, one can produce staggering lumens by using many low-power LED's as opposed to fewer high-power LED's. Assuming that the low-power LED's do not have any thermal deficiencies in their packaging and can survive long runtimes without thermal considerations, one can see them competing very effectively with power LED's so long as their unit price remains low and your "lumens per dollar" metric is similar.

For both, LED strips (if they have high flux or high efficacy) are perfect. In residential, the light source is hidden so no need for a complex fixture, for office functionality matters a lot more so again no use to build a case around it (as they do it with the LED tube).

Commercial users are driven by cost metrics and technical requirements. While crony capitalism and inertia are barriers to change, the desire to save money and get ahead will ensure that new worthwhile products gain market share.

Residential markets are different. If you've paid attention to the sound and fury about to incandescent phase-outs in numerous countries, you'll note that residential customers are extremely hesitant to change. In the US, about the only places you will routinely find linear lighting is in kitchens, garages, and - in fleeting numbers - bathrooms with vanity fixtures; elsewhere you will find tend to find fixtures with varying numbers of Edison or Candelabra sockets. Americans in particular seem to have this aversion to florescent lighting at home - perhaps it reminds them too much of work - and have voted with their dollars for incandescent (or incandescent-like) lamps.

I suspect that floros are tolerated in garages because they are utility spaces, and grudgingly in kitchens (and the rare bathroom) due to the need for daylight color rendition and prolonged operating times in the case of kitchens.

The Philips strips in question do not look like some parts of a fixture build by a 3rd party, they are just meant as light source to replace another light source, the neons.

Philips site talks extensively about upgrading existing T5 floro fixtures to LED. A chain of gas stations undergoing renovations near me has performed retrofits to their dual 4' floro fixtures (probably T12) with some flavor of strip LED; suspect they use the housing mostly as a place to stash the driver and to provide some modicum of heatsinking to the strips (thankfully they're not doing LED "tube" retrofits).

I know that the lower-power flexible-strip manufacturers have been gunning for the neon market for some time. Given that neon tops out at ~90 Lm/W (and seems to cost quite a bit to achieve that efficiency), runs at extremely high voltages, and has specialized fabrication requirements, it's no surprise that LED is a competitor. Given how the neon industry has to put their thumb on the scale with increasing pressure when doing comparisons between neon and LED, small wonder LED is eating into that market... especially with the apparent growing popularity of channel letter signs over exposed/shaped neon letters.

[wquiles]

50k hours, 100k hours lifetime was from the start the main sales pitch of LEDs and was diligently used for any fixture of any kind that had LEDs inside. For a while such claims where untrue for 99% of the products, then the serious producers started to back down from them and today only Chinese no-names still use them every day. Most fixtures from Samsung, Toshiba, Osram, Philips and many more i have seen have a claimed lifetime of 15k,25k with some exceptions in the 40k area. In all cases we need to be consider the fine print (many times not even mentioned): such lifetime is possible only if certain conditions are met, most importantly, those concerning ventilation and ambient temperature. Too bad in LED fixtures they did not put some sort of material that changes color, or smt, when exposed to a certain temperature for a prolonged interval of time. Then we could know if the fixture malfunctioned because of the heat or a flaw in the system design.

I am by the way in complete agreement with your post. I just wanted to add another data point to the particular quote above - the lack of longevity "today" is not due to the LED by itself failing. It is not due to the LED not being rated to 80-85C. Several of the Electrical Engineering magazines I subscribe to have open up several, recent LED bulb replacements and in all cases they found that the electronics were the weakest link - by far. Some other articles have seen that it is not only the use of Electrolytic capacitors a problem, but also bad thermal construction, which "will" lead to premature failure of the complete assembly, even though the LED by itself will surely last a long longer. Here is just but one of the many links about this topic - look at the photo at the end for the "drivers" - those will NOT last even close to the claimed "hours":
http://www.eetimes.com/design/smart-...semiconductors

Will

[blasterman]

A few weeks ago I had a lengthy conversation with an engineer who does independant testing for high end fixtures. He also attends all the international light fairs and has pretty good insider info due to his credentials. He claimed the biggest complaint he's hearing from higher end designers and architects doing evaluations of the latest fixtures is that manufacturers are significantly over-rating the color consistency of fixtures. While lumen values are holding, color and CCT is not. It's especially a problem in the European market where designers seem especially fussy about color temp. This may explain why there isn't much out there in the commercial sector other than generic high bay lights, and then a big jump to Cree's multi-chip lamps.

I've no reason at all to doubt this guy, and perhaps this tell us that while thermal envelopes are being pushed in passive designs and being improved on not all the phosphor technology can handle the higher thermal loads. That of course brings remote phosphor into the fray.

[bshanahan14rulz]

http://www.eetimes.com/design/smart-...semiconductors

Will

I like MuAnalysis :-) They had some interesting LED reports too back in the day

I agree, drivers these days still suck. That is why I wish the acriche idea had taken off, and I'm surprised that with all the big names coming out with lighting-class LED arrays, none of them are meant to be driven directly off of mains. I understand the whole "omg, flicker" viewpoint, but I still think it is an idea worth pursuing.

Some other ideas I've seen tossed around in this thread, I like how some of the big names are starting to rate their LEDs at higher temperatures.

On the topic of remote phosphors, the obvious solution would be to heatsink the phosphor. Just gotta find a way to do that better while not sacrificing wavelength absorption. Something perhaps like the BMW headlight idea.