200 Lumen per Watt - Cree MK-R multi die chip

http://www.cree.com/led-components-and-modules/products/xlamp/arrays-directional/xlamp-mkr

11.7v Forward Voltage up to 1.25 A driver current


Only thing wrong about it so far is that the spec sheet has the wrong PCB Solder pad size.

This looks promising...

Is anyone else seeing a lot of disparities between the announcement and the datasheet? I mean, where are they getting this 200lm/W claim? Also, why is the claimed output limit only 1769lm, where the K2 would put the output at almost 2000lm at max current.

The 200 lumens per watt is at 1 watt - somewhere between 80-100mA dependant on how the voltage drops at lower current.

If Cree is willing to stoop this low in order to derive a 200 lumens per watt headline - why didn't they do the same thing with the XM-L2 announcement a few weeks ago? My initial thought when I saw the headline is that some other manufacturer must be coming out with a similiar announcement very shortly and Cree is trying to head off the competitor's headline.

I thought 200 lm/W wasn't going to happen for a few months. oo:

These look great for general lighting. They have 5000K and 80 CRI which isn't bad at all (I might use these to retrofit some outdoor floodlights) . I wish the 90 CRI bins were available in higher CCTs. Best bins give 1040 lumens ( 127 lm/W ) @ 85°C and 1206 lumens ( 147 lm/W ) @ 25°C. If you increase current to 1.25A you'll get about 1625 lumens @ 85°C and roughly 107 lm/W. Best of all these are 4 dies in series which makes driving them much easier than a single die needing 3 or 4 amps.

slebans said:

The 200 lumens per watt is at 1 watt - somewhere between 80-100mA dependant on how the voltage drops at lower current.

If Cree is willing to stoop this low in order to derive a 200 lumens per watt headline - why didn't they do the same thing with the XM-L2 announcement a few weeks ago? My initial thought when I saw the headline is that some other manufacturer must be coming out with a similiar announcement very shortly and Cree is trying to head off the competitor's headline.

Click to expand...

Cree mentioned 186 lm/W @ 1W when they announced the XM-L2. Remember that when power LEDs first came out ~1 watt (350 mA) was more or less the industry standard. Cree in their press releases is simply showing how well their new LEDs do at this drive level. For what it's worth, they also mention the output at the maximum drive current and 85°C.

What I'm finding amazing is the best bins are well above 100 lm/W over their entire operating range.

By the way, mentioning the maximum efficacy an LED is capable of is very useful info in that it shows the maximum potential of whatever type of technology the LED is using. Any decreases from this maximum level are caused by heat and current density effects. Or put another way, if Cree wanted to, they could make an LED which gets 200 lm/W at any drive level by just increasing the total die area. Simple economics is the reason they don't.

jtr1962 said:

By the way, mentioning the maximum efficacy an LED is capable of is very useful info in that it shows the maximum potential of whatever type of technology the LED is using. Any decreases from this maximum level are caused by heat and current density effects. Or put another way, if Cree wanted to, they could make an LED which gets 200 lm/W at any drive level by just increasing the total die area. Simple economics is the reason they don't.

Click to expand...

I had not thought of efficacy in terms of total die area - I always focussed on junction temperatures.
Thanks for the info.

And boom goes the dynamite.

Reminds me of their Easy White in the G2 version. It looks promising but what really got me confused is the maximum current allowed. It states 1.25a and yet, it can run up to 15W? Anyone noticed this yet?

It's definitely a lot larger than the Easy White platform--similar to that of an MC-E except it's 7x7 and not 7x9 like the MC-E.

Tiresius said:

Reminds me of their Easy White in the G2 version. It looks promising but what really got me confused is the maximum current allowed. It states 1.25a and yet, it can run up to 15W? Anyone noticed this yet?

It's definitely a lot larger than the Easy White platform--similar to that of an MC-E except it's 7x7 and not 7x9 like the MC-E.

Click to expand...

1.25A * 11.7V = 14.625W, so yes, it is 15W.

bose301s said:

1.25A * 11.7V = 14.625W, so yes, it is 15W.

Click to expand...

So this is a low current emitter designed to produce almost 2k lumens at the emitter?

Ya.

Basic recap for those who may not know.

A die for an LED has a forward voltage of about 3V. How much current it can take is largely a function of its area. The more area it has the more current it can handle (generally). (ex. 5mm LEDs have small die and can only take 20mA while the SST-90 has a huge die and can take lots of current, 9A IIRC). In this case the LED has 4 small dies wired in series. 4x3V = 12V (close enough). Since the dies are smaller they can only handle a smaller amount of current, in this case 1.25A. This would be very similar to having 4 XR-E's wired in series. The voltage is close to 12V and the string can only take 1A (as per specs).

I hope this helps those who may not know.

Maybe they have 4 EZ1400 dies in the light. That'll make it have a similar current limit to the XP-G (1.5 A vs 1.25 A). Since there are more dies closer together, the current rating could be lower to keep heat manageable.

Looks like this l.e.d. would work great for replacing 100 watt bulbs or 23-26 watt (100 watt equivalent) florescent bulbs. 15 watts is at least a 8 watt savings over the 100 watt equivalent florescent bulb or about a 35% reduction in power consumption (probably less after circuitry losses). If these are cheap enough to use more than one per bulb, we might get closer to 150-200 lumens/watt efficiency for the bulb. I was hoping that I could get an l.e.d. flashlight with a 200+ lumen/watt l.e.d. once they became available. I don't think this l.e.d. will be going into many flashlights anytime soon. I think my best bet is to wait for higher bins of the XM-L2 as they become available.

With the voltage being near 12 volts, these may also possibly work for car l.e.d. headlight, power failure lights, and low voltage lighting.

I think that the MK-R is 4 XP-G2s wired in series and the current limit is lower than 1.5 because the dies are too close to each other. Just like in the case of the csm-360.
Also the low Vf supports this theory

Hooked on Fenix said:

Looks like this l.e.d. would work great for replacing 100 watt bulbs or 23-26 watt (100 watt equivalent) florescent bulbs. 15 watts is at least a 8 watt savings over the 100 watt equivalent florescent bulb or about a 35% reduction in power consumption (probably less after circuitry losses). If these are cheap enough to use more than one per bulb, we might get closer to 150-200 lumens/watt efficiency for the bulb.

Click to expand...

For longevity reasons these would probably be run at no more than 700 mA ( ~8 watts ). That would imply at least 2 in a 100 watt equivalent lamp. At 700 mA, two LEDs of any bin G4 or better ( or G2 if you can reduce the die temperature) would put out enough light to match 100 watts of incandescent.

If you look at the best currently available bin (J2), you can more or less match 100 watts of incandescent (1600 lumens) with any of the following:

1 LED @ 1250 mA - 107 lm/W - 10.2 watts waste heat
2 LEDs @ 550 mA - 126 lm/W - 7.9 watts waste heat
3 LEDs @ 350 mA - 136 lm/W - 7.0 watts waste heat
4 LEDs @ 250 mA - 144 lm/W - 6.3 watts waste heat
5 LEDs @ 190 mA - 153 lm/W - 5.6 watts waste heat
8 LEDs @ 115 mA - 163 lm/W - 5.0 watts waste heat
10 LEDs @ 85 mA - 178 lm/W - 4.2 watts waste heat

Note that I assume Tj of 85° in all cases and use a LER of 330 lm/W to calculate waste heat. Most like Tj would approach room temperature at lower currents, meaning efficiency would approach 200 lm/W. In any case, you're looking at 10 LEDs to reach maximum efficiency. For economic reasons this wouldn't make much sense. 2 to 4 LEDs seems to be a good compromise between cost and efficiency.

I was hoping that I could get an l.e.d. flashlight with a 200+ lumen/watt l.e.d. once they became available. I don't think this l.e.d. will be going into many flashlights anytime soon. I think my best bet is to wait for higher bins of the XM-L2 as they become available.

Click to expand...

The voltage on the XM-L2 is also better matched to the typical lithium-ion battery voltage. The MK-R would require a boost driver in most cases, sacrificing efficiency.

With the voltage being near 12 volts, these may also possibly work for car l.e.d. headlight, power failure lights, and low voltage lighting.

Click to expand...

These are definitely suited for automotive use. Because the forward voltage is only about 1 volt less than is typical for most automotive electrical systems, you can get away with a super simple linear constant current driver and still have ~90% driver efficiency.

jtr1962 said:

If you look at the best currently available bin (J2), you can more or less match 100 watts of incandescent (1600 lumens) with any of the following:

These are definitely suited for automotive use. Because the forward voltage is only about 1 volt less than is typical for most automotive electrical systems, you can get away with a super simple linear constant current driver and still have ~90% driver efficiency.

Click to expand...

Except if you are matching 100W Incan you are doing it at 3000K and the best orderable 3000K, 80CRI bin is G4.

For automotive, 11.7 is typ, 14 is max and Cree has never been real good (read amenable) to voltage based binning so I don't see this in automotive at least not a volume application without a switcher.

Semiman

Hooked on Fenix said:

With the voltage being near 12 volts, these may also possibly work for car l.e.d. headlight, power failure lights, and low voltage lighting.

Click to expand...

I always deny myself of using them on car headlights because US DOT regulations prohibits any modifications to a car's exterior lighting. It must an OEM replacement, not mods. Though, I'm sure you could do the mod for inside cargo light as the car cannot be driven with an open trunk?

Tiresius said:

I always deny myself of using them on car headlights because US DOT regulations prohibits any modifications to a car's exterior lighting. It must an OEM replacement, not mods. Though, I'm sure you could do the mod for inside cargo light as the car cannot be driven with an open trunk?

Click to expand...

I wasn't planning on modding my car headlights. I know it's not allowed. However, I think these l.e.d.s have the potential to be used in the next generation of OEM car headlights. Might make hybrid and electric cars more energy/fuel efficient. In a lot of places now, you can get a $200+ ticket for not having your car headlights on in broad daylight. This takes away the fuel savings of a hybrid car as they can't run on battery alone with the power draining headlights on. You get the increased cost of a hybrid, the increased cost of maintenance, and no savings from fuel efficiency. L.e.d. headlights have the potential to save new car owners money from using less fuel in the newest generation of hybrid and electric vehicles.