Toshiba E-CORE LED Bulb - Can't Figure it out!

[jd_oc]

I need an experts help on this one. Ok, so I've been waiting for the new Toshiba E-CORE LED bulbs to come to the US, and almost debated ordering direct from Japan where they have been available for 6+ months.

For those of you who haven't seen this here is a link:

http://www.tlt.co.jp/tlt/new/lamp/hp_led/ippan_aw8.htm

It is basically, 8.7 watts and outputs 810 lumens. That is 93 lm/W.

Just saw they are finally coming here but when I read the specs on the US Toshiba site the specs are completely different:

http://www.toshiba.com/lighting/product_display.jsp?id1=42&id2=65

It is showing 7.8 watts and outputs only 350 lumens. That is only 44.9 lm/W.

So, what is happening here? I can only think of 2 possibilities:

1) They give us americans some cheaper version.
2) The us method of measuring lumen output is different?

Anyone have any ideas? They seem nearly identical, with the model #'s being almost the same.

 

[Lynx_Arc]

almost = the key word... I cannot read any of the japanese stuff but I did see AW8N on the higher output LED which does not match the 300/350 lumen models. I would guess the 300/350 lumen models are using less LEDs driven harder while the 600-900 lumens use more LEDS driven at lower levels with perhaps more efficient LEDs to accomplish higher lumens/watt.

 

[jd_oc]

It seems so strange, though that they would differ *that* much, I mean 'half the efficacy?

For reference, the japanese model number is:
LEL-AW8L

And the US model number is:
LEL-AW8L-1/US

The exterior's look identical, so It just seems hard to believe they would debut a brand new line of lights for the US that are half as efficient as one's they've been selling in Japan for a while.

 

[LEDninja]

Probably calculated LED lumens vs OTF measured lumens.
There have been SSC-P7 flashlights sold for 3 years now with "900 lumens" proudly stenciled on the side. Recent tests by MrGman & Big-C show less than 400 OTF lumens.
The frosted dome of the bulbs absorb a lot of the light, like 1/3. If the bulb is 93 lumens per watt, the LED will have to produce 139 lumens per watt. Throw in electrical/driver losses and the LED needs to be producing 167 lumens per watt. You telling me Toshiba has an LED that can outperform the Cree XP-G by 20%?
Lets us look at it another way.
Max theoretical lumens 810.
-20% for electrical/driver loss > 648 lumens.
65% transmission > 421 lumens (I am using the 65% # for a stock Maglite, and the Maglite lens is not frosted)
72% between warm white and cool white > 303 lumens. (see next section)
Toshiba US is claiming 300 lumens for their warm white. Close enough.
Since I can estimate the US # from the Japanese # it is clearly different ways of stating things in different countries.

Also they are only bringing in the neutral & warm white to the US, not the most efficient cool white.
4sevens stated the limited run WARMS Quark & Preons are about 72% the output of the cool white versions. That means a 206 lumen cool white Quark works out to only 148 lumens for the warm white.
Turning this around:
300 lumens for the Toshiba warm white,
350 lumens for the Toshiba neutral white,
I figure 416 lumens for the Toshiba cool white.

 

[jd_oc]

Thanks LEDninja. I figured it was a difference in the way things are measured. Ok, now the question is do you think these are worth buying at around $50/each? I'll probably get a few regardless.

FYI, for all the naysayers that talk about heat dissipation, etc, limiting lumen output in LED bulbs should check this article out:

http://www.tlt.co.jp/tlt//english/info/release/release_20100412.htm

They are claiming 1690 lumens (and compare it to the AW8L, so I assume same form factor) and an efficacy of 120lm/W. Also has a lighting angle of 260 degrees. A light like this if priced right could really bring LEDs into the mainstream. Of course we just have to wait until they are available

 

[blasterman]

and an efficacy of 120lm/W. Also has a lighting angle of 260 degrees. A light like this if priced right could really bring LEDs into the mainstream.

120lm/W is higher than what Cree is getting with top bins, so unless Toshiba is sitting on some miracle LED technology that their share holder might want to know about.....

Or, this are just more low CRI cool-white emitter propoganda. Cree is also the most notorious for this.....

Also, Asian countries where a lot of this stuff is made have pretty lax advertising laws and can claim what they want. Consumers buy it, say "say it sucks", then move on.

 

[jtr1962]

Thanks LEDninja. I figured it was a difference in the way things are measured. Ok, now the question is do you think these are worth buying at around $50/each? I'll probably get a few regardless.

FYI, for all the naysayers that talk about heat dissipation, etc, limiting lumen output in LED bulbs should check this article out:

http://www.tlt.co.jp/tlt//english/info/release/release_20100412.htm

They are claiming 1690 lumens (and compare it to the AW8L, so I assume same form factor) and an efficacy of 120lm/W. Also has a lighting angle of 260 degrees. A light like this if priced right could really bring LEDs into the mainstream. Of course we just have to wait until they are available

You're misreading a couple of things. First off, there is no mention of power consumption for the 1690 lumen AW8L lamp. In all likelihood they're using some form of active cooling, and overall efficiency is probably well under 100 lm/W. I could squeeze probably 3K lumens from a screw-base form factor using the best available LEDs and active cooling if I wanted to. However, it certainly won't be cheap, and I doubt it would exceed 120 lm/W.

Second, the 120 lm/W refers to what is called an "LED base lamp". They go on to say: Base lights mount arrays of LEDs in a line, combining their output to achieve effective brightness with comfortable light. They integrate highly efficient LED and power circuit that achieve a brightness of 120lm/W, surpassing the efficiency of fluorescent lights.

In short, this is probably what we at CPF call a "purpose-built LED fixture". Hmm, haven't we here at CPF ( at least a small minority of us anyway ) been saying all along ditching the antiquated screw base form factor is what's needed to get these super-high efficiencies?

BTW, I've little doubt we will exceed 1500 lumens, even 2000 lumens, in a screw-base form factor without active cooling, but that won't occur until the raw efficiency of LEDs exceeds 200 lm/W. Cool-white is already there, at least in the lab. Neutral and warm white has a way to go, although I suspect that for various reasons the popularity of these tints is on the way down. They already use 5000K for residential lighting in Japan, have for a while now. We'll probably follow suit, both because it's more efficient, and also generally easier on the eyes. We just need to get the CRI of cool-white up to something acceptable. Actually, 5000K is closer to a cool tint of neutral white, but the point remains. It makes no sense to try and imitate incandescent lamps with LEDs. You lose a ton of efficiency. You can still get decent CRI by increasing the color temperature without sacrificing as much efficiency. Besides, 2700K and 3000K LEDs also ( IMO ) look awful, reminiscent of HPS lights. I've been seeing a lot more homes lit with cooler tint CFLs lately. Apparently once people are no longer wedded to incandescent lamps, they're discovering they prefer a whiter tint. Japan figured this out years ago. It's a shame makers of LED lamps aren't catching on to this fact yet. All you have is a very warm white, or a very cool white. Generally there is nothing in between.

 

[LEDninja]

Ok, now the question is do you think these are worth buying at around $50/each? I'll probably get a few regardless.

EarthLED just announced the Zetalux 2 bulbs 6W 300L warm white 400L cool white at $24.99 each + shipping. Lifetime of the bulb is a low 25,000 hours (that is still 3 years 24/7 or 12 years 6 hours a day)

EarthLED also announced the Zetalux 2-pro 7W 450L warm 550 cool at $34.99 + shipping. I am very suspicious of the lumen numbers of this one as they are claiming 50% more lumens for a 16% increase in power input.

These bulbs use a less familiar Cree MX-6 LED.
http://www.cree.com/products/xlamp_mx6.asp

 

[blasterman]

In short, this is probably what we at CPF call a "purpose-built LED fixture". Hmm, haven't we here at CPF ( at least a small minority of us anyway ) been saying all along ditching the antiquated screw base form factor is what's needed to get these super-high efficiencies?

Actually, what I'm reading between the lines is they're using the same smoke and mirrors as were seeing in the reef tank industry. They're using a smaller projection angle due to optics and claiming more efficiency. When in doubt, we just need to look up the base efficacy of the emitter being used, and the 'fig leaf pops off' so to speak. If the emitter being used can't produce more than 100lm/w, then the fixture won't either.

although I suspect that for various reasons the popularity of these tints is on the way down. They already use 5000K for residential lighting in Japan, have for a while now. We'll probably follow suit, both because it's more efficient, and also generally easier on the eyes.

Again, we may agree to disagree on this, but I don't see current mediocre CRI cool-white LED technology getting anywhere other than:

(1) Getting the flashlight guys sexually aroused
(2) More reason for Cree to put out press releases to get their uneducated shareholders sexually aroused :tinfoil:

From what I'm seeing it's cool-white LED technology that's failing in the retail market, and I'm seeing an explosion of better quality retrofits all using neutral and warm-white emitters. Initially cool-white LEDs were the first exposure most people had to working under LED, and that has lead to a big backlash against the industry. Initially a lot of people bought 6500k CFLs because they had a lot of flashy marketing, then took them down when their spouse started making threats.

People work all day at the office under 4000-5000k light meant to stimulate productivity, and when I get home I want to relax.

We both agree that a high CCT, high CRI emitter are needed, but it doesn't look like the industry seems that interested. Once you mix in enough red and amber to get a +90 CRI your lumen efficiency gets murdered anyways. Plus, what the industry calls 'warm white' in terms of LED is usually around 3500k with some pink in it. I've never seen an LED with the same color as an incan.

Off topic jtr, but are you using a spectrometer for your LED tests? I have somebody building me a custom 10watt 440nm LED, and I need somebody to test it. I'll take care of postage.

 

[blasterman]

I am very suspicious of the lumen numbers of this one as they are claiming 50% more lumens for a 16% increase in power input.

Hmm, hasn't the lifespan of the newer bulbs bulb gone down though? This would tell me they're just beating the emitters harder. The price is what's really goof because if they're using Crees their profit margin is pretty darn small.

What's frustrating is you could get a +800lumen Bridgelux emitter in a PAR 38 format, design a decent radial heat sink, and with the same active cooling get +70K hours out of the thing. You'd make money selling it for less than $50. So, it looks like EarthLED has given in to disposable bulb mentality for profit.

 

[jtr1962]

From what I'm seeing it's cool-white LED technology that's failing in the retail market, and I'm seeing an explosion of better quality retrofits all using neutral and warm-white emitters. Initially cool-white LEDs were the first exposure most people had to working under LED, and that has lead to a big backlash against the industry.

Don't forget though you're talking about 8000K-10000K "showerhead" lamps using cheap LEDs. This is quite different from the 5000K or so I'm talking about.

Initially a lot of people bought 6500k CFLs because they had a lot of flashy marketing, then took them down when their spouse started making threats.

Same problem as the aforementioned LED lamps. Lack of quality control or standards meant so-called 6500K was either 8000K in reality, and/or so far off the blackbody curve that it was strongly blue-tinted even if it was 6500K according to a spectrometer. True 6500K, mostly from linear tubes, isn't that bad. CFL quality control though is atrocious compared to linear tubes. Sure, 6500K is a big jump from 2700K incandescent, so they might be some getting used to. In the end though if you're used to daylight it's not as bad as you make it out to be. Of course, if people won't even give it a chance...

People work all day at the office under 4000-5000k light meant to stimulate productivity, and when I get home I want to relax.

My take on this is a little different. If I'm home and need anything more than a nightlight on, it generally means I'm doing some task requiring artificial lighting. As such, I want light conducive to seeing well, which is generally higher CCT. If I'm relaxing, that either means I'm watching TV or in bed. In both cases I simply just turn the lights off. No need in my book for low CCT lighting because some say it's "reflexing". Frankly, I find the opposite to be true. When I can't see well because colors are distorted to hell, I find it stressful. I certainly understand the need to have decent CRI ( or whatever new metric eventually is used in its place ), but good CRI and higher CCT aren't mutually exclusive. The big problem with low CCT lighting is the lack of a true white point, plain and simple. You can't tell looking at a white object whether it's white or yellow.

We both agree that a high CCT, high CRI emitter are needed, but it doesn't look like the industry seems that interested. Once you mix in enough red and amber to get a +90 CRI your lumen efficiency gets murdered anyways.

Look at this post. High-CRI cool white gives you about a 30% lumen decrease relative to low-CRI cool white. The penalty for high-CRI gets worse as the color temperature decreases. Also note how much efficiency gets murdered reducing CCT even without increasing CRI. It's a shame the industry isn't much more interested in high CRI and high CCT. In a sense, they could have their cake and eat it too, at least compared to the alternatives. I'm still waiting on a new metric to measure color accuracy which will hopefully take both CCT and spectrum into account. That might be the push needed to make high-CRI, high CCT LEDs ( these would be more efficient, and likely also do better under the new system compared to high CRI neutral or warm emitters ).

Plus, what the industry calls 'warm white' in terms of LED is usually around 3500k with some pink in it. I've never seen an LED with the same color as an incan.

SSC has 3000K high-CRI LEDs. Also, I think GE has aimed for 2700K in their latest LED screw-base lamp discussed in another thread. If the LED industry had stopped at 3500K then frankly I wouldn't be as vocal in my complaints as I am. 3500K still has a steep efficiency penalty, but asthetically at least it's tolerable to me ( barely ). 3000K and especially 2700K just isn't. The problem isn't where I live as I can likely find or make any type of light I like, but rather if I visit others ( my brother for example ). Can't standing sitting under the damned incandescents he uses. You're talking major headaches.

Off topic jtr, but are you using a spectrometer for your LED tests? I have somebody building me a custom 10watt 440nm LED, and I need somebody to test it. I'll take care of postage.

I'm just using a regular light meter unfortunately. I'd love to get a spectrometer, or at least a color temperature meter so I could do better than eyeballing the LEDs I test, but it's well out of my budget.

 

[sendjerry]

There are few differences between these 2 models
1. The model with high efficacy only sold in Japanese market where there is no safety codes required, maybe soon in the future but not now. The models sold in US market must pass the safety codes of FCC or UL. In order to pass the safety requirement, there are some components in the model for Japanese market must be replaced but that makes power efficacy drop. That [email protected] model has EMS issue and its power factor is only around 0.6. I am a LED light bulb designer. We have the same problem. In order to pass the safety codes like FCC or CE (for European market), we have to push the power consumption to 10 watts to generate 800 lm but it passes the FCC and CE. This is why the efficacy is so different in these two models.
2. Another difference between these 2 models is that the 810lm@ 8.7W model uses COB type emitters (66 small emitters) and the one with is SMT type (7 large emitters). Heat dissipation is easier for COB type.
3. The body of the model is larger and heavier. It weights 160g instead of 135. Its heat sink is big and more powerful to dissipate heat which also makes the power efficacy higher.

 

[jeffosborne]

Hi sendjerry, and welcome to CPF!
Thanks for your insightful comments. And how exciting to hear you are a light designer! Where are you located? We would all be happy to hear some 'insider' stories you might have, so be sure to share. For example, what are you working on now?

Cheers,
Jeff