'Most energy-efficient' LED light revealed by Philips

Nice, but two years before we see it available??? Sounds rather experimental at this point. At 200lm/W, I’d expect someone else to match or surpass it by the time they release it to the market. I guess they have two years to get the price to a reasonable level.
Companies have been working on LED-based retrofit fluorescent tubes for years and silly old-school regulations (tailored to traditional technology) have been hindering their release and driving up costs. I’m finally hearing more about their release soon in the US market.
The kind I’m referring to can be inserted in place of a fluorescent tube without changing the ballast. No electrician required.

sometimes i think plug in retrofits can be a mixed bag, they can facilitate adoption, but they can also limit what new technology is capable of accomplishing

I know a few people that put in fancy LED fixtures in their new houses / remodels. All sorts of wonderful promises. A couple years later they needed parts and voila… no longer available. Stick with fixtures that use good old fashioned retrofit bulbs.

if something better does come out and no one adopts it, then we are voluntarily remaining in the dark ages.
but you are correct, that kind of thing should not happen yet it does with alarming frequency

And if your house is lit with fixtures that can’t be fixed… you will remain in the dark.

i agree, so how does one promote the potential of the better the technology without ending up in the dark?

Interesting and very plausible,their newest luxeon T already smashed its competitor xp-g2 in every aspect:efficiency,CRI,voltage drop,thermal resistance.It’s even more efficient than xm-l2 u2 at same current densitiy and it’s 70cri min vs 65cri typ. for xm-l2 u2.

I think of BLF the main thing is how many amps can it pull, not how efficiency it is.

What's the typical luminous flux at 1000mA?

Interesting, but strip lights are ugly as sin. :Sp

I made a little table in which you can see, I called it "thermal performance" of popular leds.

This number says what would be thermal resistance of LED for each 1mm2 die area.The lower,the better,because

thermal resistance is the main limitation for overdriving LED.

As you can see,xp-g2 can be driven harder than xm-l2 (higher mA/mm2) because xp-g2 thermal performance is 8 C/(W*mm2) vs. xm-l2 10 C/(W*mm2).

And luxeon T has 6 C/(W*mm2) !

In first table you can see typical(so not maximum bin) values for flux and efficacy. 143lm/w@350mA , 130lm/w@700mA , 122lm/w@1000mA , very little drop for 3x increase in current! From this and very low thermal resistance of 3C/W you can conclude this LED can be driven harder than any other LED.

Another advantage over xp-g2 is "wireless" die connection,so no extra voltage drop at those tiny wires at high currents,and generally lowest Vf among all white leds.

Part
Number
CCT (K) Min CRI Minimum Luminous
Flux (lm)
Typical Luminous Flux (lm) Typical Efficacy (lm/W)
700 mA 350 mA 700 mA 1000 mA 350 mA 700 mA 1000 mA
LXH7-FW30 3000K 70 200 117 218 298 123 111 104
LXH7-FW40 4000K 70 210 133 249 340 140 127 119
LXH7-FW57 5700K 70 220 136 255 348 143 130 122
LXH8-FW27-Y 2700K 80 170 100 186 254 105 95 89
LXH8-FW30-Y 3000K 80 180 106 197 269 112 101 94
LXH8-FW35-Y 3500K 80 190 112 208 278 118 106 97
LXH8-FW40-Y 4000K 80 190 114 212 288 120 108 101
LXH8-FW50-Y 5000K 80 190 120 222 300 127 113 105
LXH9-FW30-Y 3000K 90 130 86 160 216 91 82 76

LED

DIE SURFACE (mm2)

THERMAL RESISTANCE (C/W)

THERMAL PERFORMANCE (C/W)*mm2

NICHIA219

2

7

14

SST-50

5

2.45

12,25

MC-E

4x1

3

12

Rebel ES

2

6

12

XM-L(2)

4

2.5

10

Luxeon M

4x2

1.25

10

XP-E(2)

1

9

9

Rebel

1

9

9

XP-G(2)

2

4

8

XR-E

1

8

8

Luxeon T

2

3

6

SST-90

9

0.64

5,76

The Luxeon T looks like a really attractive led I must say! It will not convince the lumen freaks but in every other aspect I like it for a flashlight. Two concerns: 1) will it ever be available other than from Mouser or Digikey (I can't afford those shipping costs), 2) the solder pad lay-out is completely different from Cree/Nichia, and looks more difficult to DIY reflow, so it can only be used presoldered on stars (and those will probably not be Sinkpads or the like where you want such a low thermal resistance led)

I see what you mean led4power, however the XP thermal path is 3 times bigger than the one on the Luxeon T LED, so I am not sure show that overdrive would work in the end compared to over-driving the XP-G2 which is close to the Luxeon T in thermal performance and just a tiny bit more efficient at 1000mA.

XP thermal pad isn't 3x bigger,it's about 2 times bigger,but bigger thermal pad doesn't really help in reducing thermal resistance because of relatively small thermal conductivity and thickness of substrate.Thermal pad area directly behind the die (cca 1.4x1.4mm fro 2mm2 die) transfers most of heat.

Luxeon T : 3C/W

XP-G2 : 4C/W

3 as a number is maybe close to 4,but 3C/W vs. 4C/W is a BIG difference,it means that you can drive lux T 33.3% harder!

In reality,you should push lux T even harder than 33.3%,because of lower Vf(at 3+ amps difference in Vf should be even higher because lux T doesn't have gold wires,die is directly connected to a-k pads),and lower Vf means less power=V*I,so you can push more amps to equal xp-g2 in generated heat.

How ?

xp-g2 at 1000mA,85C ::

338lm,

Vf=3V

eff=338/(1*3)=112.7lm/W

lux T at 1000mA,85C:

348lm,

Vf=2.86V

eff=348/(1*2.86)=121.7lm/W

Typical(not max bin!) luxeon T is more efficient than top binned R5 xp-g2,it's even more efficient than top binned XM-L2 U2 at 500ma/mm2 (1Amp lux T, 2 Amps U2),and all that with 70CRi minimum,5700 kelvin, versus 70cri typical for xp-g2,and 65cri typical for xm-l2.

Ok, I like it.

Regarding the thermal pads, it's 1.3 * 3.3 vs 1.4*1.7, I think got the numbers right, so it's only 2 times indeed.

You deal in LEDs?

led4power,
Thanks for bringing the Luxeon T to my attention. It’s got potential and wasn’t even on my radar.
-Slim

You deal in LEDs?

Can you reconfigure the question?(my english sucks :* ).

It does not suck. What I meant are you a LED dealer?

No,I just love LED technology.

Actually it seems like the thermal pad is 1.25*1.45mm for the Luxeon T if i look at the mechanical dimensions and not at solder pad design.