The new XP-G: another big announcement from CREE

[Linger]

Nice find Marduke.
He's been hinting something was happening today; their interview was scheduled for release.

 

[Nedtheshred]

Glad I have CREE stock.
These are awesome!!!

http://www.globenewswire.com/newsroom/news.html?d=174391

Ned

 

[saabluster]

Glad I have CREE stock.
These are awesome!!!

http://www.globenewswire.com/newsroom/news.html?d=174391

Ned

Here is a notable quote.

"we have begun offering limited samples of an S2 flux bin, providing up to 400 lumens at 1A. We target availability of XP-G neutral and warm white LED samples by end of calendar 2009."

 

[rav]

Check this out! :huh::huh::huh:

http://www.cree.com/xlamp/sample_xpg.asp


:wow:

 

[saabluster]

Check this out! :huh::huh::huh:

http://www.cree.com/xlamp/sample_xpg.asp


:wow:

That is for industry use not the general public.:poke:

 

[uplite]

It would be interesting to see a graph lm/w/mm^2 v. Cur. for XP-E R2 and XP-G R5 to compare difference in technology.

According to posts in this forum, the XPE die is 0.9mm x 0.9mm and the XPG die is 1.4mm x 1.4mm. Now that the XPGs are public...can someone confirm or deny these numbers?

If you divide marduke's efficacy numbers (lm/W) by 0.81mm2 (XPE) and 1.96mm2 (XPG):

350mA:
XP-E R2 ... 128 lm/W/mm2
XP-G R5 ... 67 lm/W/mm2

700mA:
XP-E R2 ... 106 lm/W/mm2
XP-G R5 ... 58 lm/W/mm2

1000mA:
XP-E R2 ... 90 lm/W/mm2
XP-G R5 ... 52 lm/W/mm2


We might speculate that Cree has left plenty of headroom for S and T bins with the XP-G die. But don't quote me on that. ;)

-Jeff

 

[mudman cj]

Here is a notable quote.

"we have begun offering limited samples of an S2 flux bin, providing up to 400 lumens at 1A. We target availability of XP-G neutral and warm white LED samples by end of calendar 2009."

:clap::party:

 

[rav]

. . .

 

[jtr1962]

According to posts in this forum, the XPE die is 0.9mm x 0.9mm and the XPG die is 1.4mm x 1.4mm. Now that the XPGs are public...can someone confirm or deny these numbers?

If you divide marduke's efficacy numbers (lm/W) by 0.81mm2 (XPE) and 1.96mm2 (XPG):

350mA:
XP-E R2 ... 128 lm/W/mm2
XP-G R5 ... 67 lm/W/mm2

700mA:
XP-E R2 ... 106 lm/W/mm2
XP-G R5 ... 58 lm/W/mm2

1000mA:
XP-E R2 ... 90 lm/W/mm2
XP-G R5 ... 52 lm/W/mm2

A better metric for comparison is lumens per mm² at any given current density. For example, at 350 mA the XP-E R2 (or any bin XP-E using the same die size) has a current density of 432 mA/mm². If we assume middle of the bin lumens (118), then we get 118/0.81 or 145.6 lumens/mm² at a current density of 432 mA/mm².

Now let's look at the XP-G R5. In order to obtain the same current density, you need to drive the LED at 847 mA. According to the data sheet, output at 847 mA would be roughly 315 lumens (again assuming middle of the bin lumens at 350 mA). So this comes out to 315/1.96, or 160.7 lumens/mm² at the same current density where the XR-E R2 only managed 145.6 lumens/mm².

In short, this is a significant improvement in output for any given amount of current density. I'll guess at higher current densities, the XP-G has an even greater edge.

I have some R4s on order from Cutter (I think Cutter is shipping R5s for whoever ordered an R4 because that's what they got in first). Whether it's R4 or R5, rest assured I'll be doing tests on it along with plenty of comparisons to the XR-E R2 (which I already tested).

 

[uplite]

A better metric for comparison is lumens per mm² at any given current density. For example, at 350 mA the XP-E R2 (or any bin XP-E using the same die size) has a current density of 432 mA/mm². If we assume middle of the bin lumens (118), then we get 118/0.81 or 145.6 lumens/mm² at a current density of 432 mA/mm².

Now let's look at the XP-G R5. In order to obtain the same current density, you need to drive the LED at 847 mA. According to the data sheet, output at 847 mA would be roughly 315 lumens (again assuming middle of the bin lumens at 350 mA). So this comes out to 315/1.96, or 160.7 lumens/mm² at the same current density where the XR-E R2 only managed 145.6 lumens/mm².

That is a very interesting angle! :)

Remember to also consider thermal management. An emitter driven at 850mA generates about 2x as much heat as one driven at 350mA. Higher temp = lower flux...which may kill any theoretical gains in output. And higher temp = lower VF = higher current = higher runaway potential. If you want to run high currents, it's all about the system thermal design. :thumbsup:

-Jeff

 

[Marduke]

That is a very interesting angle! :)

Not really an "angle", but rather industry standard.

 

[AlexD]

A better metric for comparison is lumens per mmІ at any given current density.

Here Vf difference isn't taken into account. When i offered "a graph lm/w/mm^2 v. Cur." i meant at the same current density not just dividing by die area.

 

[Marduke]

Here Vf difference isn't taken into account. When i offered "a graph lm/w/mm^2 v. Cur." i meant at the same current density not just dividing by die area.

The center bin efficancy at those values are nearly identical.

 

[ubetit]

I already have a bunch of useless knowledge (says the wife) in my skull and now you guys are filling it with more. :)

 

[AlexD]

The center bin efficancy at those values are nearly identical.

You mean that Vf of XP-E R2 and XP-G R5 at the same current density are nearly identical ?

 

[uplite]

You mean that Vf of XP-E R2 and XP-G R5 at the same current density are nearly identical ?

:popcorn:

 

[znomit]

I thought the smaller die was only in use on some XR-Es. XR-Es are still in limited release in R2(check the cree datasheet, no R2 listed), hence the output comparisons with the XP-G are done with XR-E Q5s. What bin are the smaller die XR-Es?

 

[Marduke]

You mean that Vf of XP-E R2 and XP-G R5 at the same current density are nearly identical ?

Efficacy, not necessairly Vf (don't have the raw data in front of me). The flux vs I plays a role too.

However, I was mistaken above. I was looking at the R4 curve and not the R5 curve. At the same current density, the XP-G R5 is ~5 lm/W more than the XP-E R2.

 

[uplite]

I thought the smaller die was only in use on some XR-Es. XR-Es are still in limited release in R2(check the cree datasheet, no R2 listed), hence the output comparisons with the XP-G are done with XR-E Q5s. What bin are the smaller die XR-Es?

IIRC, the recent XR-Es and XP-Es have the same die size, about 40% of the area of an XP-G. Disclaimer: Those measurements come from other folks in this forum. :)

The much larger XP-G die will give you more lumens, naturally, even if efficacy/area is lower.

As jtr1962 mentioned, IN THEORY, you can drive an XP-G R5 at high current to get similar efficacy/current density as an XP-E R2. However...bear in mind that heat is a REAL-WORLD factor at higher currents. Cree quotes most of their datasheet specs for Tj = 25 degrees C... :laughing:

-Jeff

 

[TorchBoy]

It's daylight again down under. Is http://www.philipslumileds.com/pdfs/DS23.PDF what you need Marduke?