Anatomy of Luxeon 3W Star vs. 1W Star

Just receive 3W Star from ElektroLeumen. Thanks Wayne!

Very interesting, the 1W star uses aluminum-core PCB exclusively. The 3W star uses a standard PCB with a hole in the middle attached to aluminum star heatsink. The 3W emitter leads are attached to PCB. The emitter slug passes through PCB hole stopping flush to back of PCB. The PCB along with the emitter slug is then glued to aluminum star heatsink. This enables emitter slug to be directly glued to aluminum star heatsink.

I measured from bottom of star to top of black ring:
1W HD Star = 0.167"
3W HD Star = 0.162"
Difference = 0.005"

Luxeon data sheet shows:
1W aluminum-core PCB = 1.6mm or 0.063"
3W aluminum heatsink = 1.5mm or 0.059"
Difference = 0.004"

Is 1/1000" possible discrepancy responsible for improved heatsinking of 3W?
Is new process for 3W cheaper then old method?
Never seen a 5W star. Do they have a heatsink design similar to 3W or 1W?

I need someone to measure the black rings of bare emitters, 3W HD and 1W HD from bottom of slug to top of black ring to verify if there is indeed a discrepancy. The black rings could have been redesigned and have different dimensions.

Thanks,
Tim

3W's stars are setup like 5W stars. So a 5W star will be like the 3W your're seeing.

Thanks IsaacHayes, I appreciate your feedback.

I am surprised no one else has measurements to share, answers or comments on my questions! I will call Lumileds tomorrow and try to get the measurements myself. I will then post for all to see. Lumileds datasheet for 1W, 3W and 5W HD emitters show 5.86mm or 0.230" from slug to top of HD. It does not have to top of black ring measurement. If they are all identical then thinner gluing is probably responsible in improved heatsinking. I don't trust Lumileds datasheet as you will see later.

I also measured from bottom of star to top of lens:
1W HD Star = 0.295"
3W HD Star = 0.287"
Difference = 0.008"
Lumileds datasheet says: 7.5mm or 0.295" for both the 1w and 3w Stars. As you can see that's not what I got!

More observations I made:
1W HD Star: uses narrow leads and the wider leads are sniped off
3W HD Star: uses wider leads and the narrow leads are sniped off.
This may also help heat management.

As most of us know, Lumileds improved the phosphor coating when they went from LD to HD. They use less phosphor and only coat the die on the HD (lambertian). Previously LD (Batwing) had a big glob of phosphor. Lumileds does not even offer LD on 3 or 5w versions. Is it because LD is inferior? Well, from my observations I cannot see any visible improvements here. The 1W HD looks like the 3W HD in the phosphor department. I know Lumileds is claiming color and heat improvements etc. I just can't see it. In fact everything under the lens look identical.

I'm sure everyone has noticed improved luminous flux and efficiency of the 1w over time. Maybe the 1w is maxed out and the 3w is positioned to evolve.

I hope these observations are helpful.

-Tim

Hello again,

I have been studying the datasheets. The 1w HD Star has extra dielectric (epoxy) layer 0.10mm and 35um copper layer thickness that the 3w doesn't. This is why 3w star aluminum heatsink measures 1.5mm and 1w star MCPCB is 1.6mm. Because 3w is bonded directly to heatsink (bypassing dielectric epoxy layer) it should have improved heat management.

How did Lumileds achieve better heat management for the 3w bare emitters? I haven't checked, but maybe Lumileds requires designers to use a lower thermal resistance on the 3w bonding then the 1w. In this way they could overdrive a 1w to some degree.

-Tim

Curiously, both the 1W Star and Lux III Star datasheets list the same junction to board thermal resistance of 17C/W. This is despite the bare emitter Lux III datasheet listing a lower junction to slug resistance than the 1W. It doesn't make sense to me.

[ QUOTE ]
startravel said:
I'm sure everyone has noticed improved luminous flux and efficiency of the 1w over time. Maybe the 1w is maxed out and the 3w is positioned to evolve.

-Tim


[/ QUOTE ]

Tim, you make an excellent point here. The 1W I think is approaching its potential and driving to higher current levels is required to continue the evolution to higher output from a single die. Keep us posted and keep up the good work.

CM