Cree 100 lm/W LED

[Victor Roberts]

According to the following press release Cree has demonstrated a 100
lm/W LED with what they call 50% wall plug efficiency.

http://www.cree.com/News/news214.asp

However, I can't find any data on the Cree Web site that supports this
claim. They _could_ be talking about a green LED, but then the 50%
number would be wrong. The eye sensitivity in the green is abut 683
lm/W so 50% efficiency would mean 341 lm/W, not 100 lm/W. The
combination of 100 lm/W and 50% efficiency implies white light, since
200 lm/W is often used as the maximum efficiency of a "white" light
source.

Does anyone have more information about Cree's claim?

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

 

[Clive Mitchell]

According to the following press release Cree has demonstrated a 100
lm/W LED with what they call 50% wall plug efficiency.

"50% wall plug efficiency" Is that like 1000W Peak Music Power Output?

 

[Victor Roberts]

"50% wall plug efficiency" Is that like 1000W Peak Music Power Output?

It is not clear what they mean here, but the term "wall plug efficacy"
is well understood to mean the efficacy of the light source and its
associated ballast or power supply.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

 

[Don Klipstein]

According to the following press release Cree has demonstrated a 100
lm/W LED with what they call 50% wall plug efficiency.

http://www.cree.com/News/news214.asp

However, I can't find any data on the Cree Web site that supports this
claim. They _could_ be talking about a green LED, but then the 50%
number would be wrong. The eye sensitivity in the green is abut 683
lm/W so 50% efficiency would mean 341 lm/W, not 100 lm/W. The
combination of 100 lm/W and 50% efficiency implies white light, since
200 lm/W is often used as the maximum efficiency of a "white" light
source.

Does anyone have more information about Cree's claim?

My guess this was a demonstration prototype made with a chip that is
either not yet in production or was a production one where variations due
to tolerances went favorable. However, I cannot yet rule out unusual
operating conditions.

The best chips in any datasheets that I found so far in Cree's website
produce 24 mW of blue light from 68 mW of electrical input if I remember
correctly - that's 35% "wallplug efficiency".

- Don Klipstein ([email protected])

 

[Victor Roberts]

My guess this was a demonstration prototype made with a chip that is
either not yet in production or was a production one where variations due
to tolerances went favorable. However, I cannot yet rule out unusual
operating conditions.

The best chips in any datasheets that I found so far in Cree's website
produce 24 mW of blue light from 68 mW of electrical input if I remember
correctly - that's 35% "wallplug efficiency".

- Don Klipstein ([email protected])

If that is 68 mW into the LED I would argue that is NOT wallplug
efficiency since the power supply losses are not included. However, as
with "full spectrum" wall plug efficiency is not a term defined by any
standards organization so I guess they are allowed to make up any
definition that supports their advertising objectives.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.

 

[Daniel J. Stern]

"50% wall plug efficiency" Is that like 1000W Peak Music Power Output?

Perhaps they're measuring in internet inches.


-DS, thinking of those Chinese "3 million candlepower" flashlight lanterns
you can buy for $18 from Wal-Mart, that use a 6v 6w bulb...

 

[Clive Mitchell]

-DS, thinking of those Chinese "3 million candlepower" flashlight
lanterns you can buy for $18 from Wal-Mart, that use a 6v 6w bulb...

I measured the optical output of one of those lamps and it actually
exceeded 3 million candlepower. Mind you, it didn't last too long
across 240V.

 

[Don Klipstein]

If that is 68 mW into the LED I would argue that is NOT wallplug
efficiency since the power supply losses are not included. However, as
with "full spectrum" wall plug efficiency is not a term defined by any
standards organization so I guess they are allowed to make up any
definition that supports their advertising objectives.

I agree there. I think they should use "conversion efficiency". Then
again, they don't say "wallplug" on the chip datasheet. But then again
then again I am guessing where they do use the word "wallplug" they either
mean conversion efficiency or else have the most efficient ballast they
can get their hands on.

But if an LED adoes produce 100 lumens of white light per watt of
electricity that gets delivered to the LED, that's still awfully good and
I wonder when I get to get one!

- Don Klipstein ([email protected])

 

[Don Klipstein]

Perhaps they're measuring in internet inches.


-DS, thinking of those Chinese "3 million candlepower" flashlight lanterns
you can buy for $18 from Wal-Mart, that use a 6v 6w bulb...

I have noticed one brand of lower cost ones with higher brightness
numbers and often more than one model on the shelf, and that lately:

They no longer use the word "candlepower" with the big number, but
"power series". At least at Target I noticed this.

- Don Klipstein ([email protected])

 

[Victor Roberts]

I agree there. I think they should use "conversion efficiency". Then
again, they don't say "wallplug" on the chip datasheet. But then again
then again I am guessing where they do use the word "wallplug" they either
mean conversion efficiency or else have the most efficient ballast they
can get their hands on.

But if an LED adoes produce 100 lumens of white light per watt of
electricity that gets delivered to the LED, that's still awfully good and
I wonder when I get to get one!

You may have noticed that the e-mail address listed for more info on
the press release has no domain name. I plan to call them on Monday
and try to get more info. I don't think this is real, but would like
to figure out what combination of number they put together to make
this claim.

--
Vic Roberts
http://www.RobertsResearchInc.com
To reply via e-mail:
replace xxx with vdr in the Reply to: address
or use e-mail address listed at the Web site.