1W chips in 5W packages?

[Bandgap]

I notice 5W and 1W Luxeon Star assemblies are different, with a lower thermal resistance on the 5W models.

Does anyone know if there will be a '1W' chip available mounted in the thermally better '5W' Star package?

Steve

 

[Doug S]

Originally posted by Bandgap:
I notice 5W and 1W Luxeon Star assemblies are different, with a lower thermal resistance on the 5W models.

Does anyone know if there will be a '1W' chip available mounted in the thermally better '5W' Star package?

Steve

<font size="2" face="Verdana, Arial">Steve, the 5W package is not thermally better in the way that you are thinking. The published lower thermal resistance is a consequence of having 4X the die area in contact with the slug. A single 1W die in the 5W package would not give you the published 5W thermal resistance value. I am currently trying to explain the concepts involved in the following thread:
http://www.candlepowerforums.com/cgi-bin/ultimatebb.cgi?ubb=get_topic;f=14;t=000880

 

[Bandgap]

Funnily enough, it was reading Doug S's fine thermal articles that reminded me to ask the question.

I am sure than having 4 times the die area cuts chip to slug thermal resistance by 4 (although there must be some electrical isolation here in the 5W device)...

...but the packages are definitely different.
My 5W white has a thin FR4 layer and the central die holder goes down to the aluminium through a clearence hole - as far as I remember.
My 1W whites appear to sit straight on the FR4.
And the cross sections in the thermal data sheets are also different - although this could be diagrammatic license.

If the differences are not for thermal reasons, does anyone know what they are for?

Steve

 

[Doug S]

Originally posted by Bandgap:
Funnily enough, it was reading Doug S's fine thermal articles that reminded me to ask the question.

I am sure than having 4 times the die area cuts chip to slug thermal resistance by 4 (although there must be some electrical isolation here in the 5W device)...

...but the packages are definitely different.
My 5W white has a thin FR4 layer and the central die holder goes down to the aluminium through a clearence hole - as far as I remember.
My 1W whites appear to sit straight on the FR4.
And the cross sections in the thermal data sheets are also different - although this could be diagrammatic license.

If the differences are not for thermal reasons, does anyone know what they are for?

Steve

<font size="2" face="Verdana, Arial">Steve, I am interested in your observation that the 5W emitters are mounted differently on the MCPCB than the 1W emitters [if I am interpreting you correctly]. I have never removed a 5W from a Star board. Perhaps someone else can jump in and help us here. I have noted that comparing the 1W and 5W datasheets, the thermal resistances of the bare emitter vs the Star configuration differ by 2C/W for the 1W and 3C/W for the 5W implying a higher slug to board resistance for the 5W. Since the values are only specified to whole interger values, I had always assumed that the apparent difference may just reflect "rounding" error.

 

[MicroE]

Originally posted by Bandgap:
....5W white has a thin FR4 layer and the central die holder goes down to the aluminium through a clearence hole....

Steve

<font size="2" face="Verdana, Arial">Doug---This is correct. The 5W slug sits directly on a star-shaped sheet of aluminum. There is a thin epoxy layer to hold it in place, as you would expect.---Marc

 

[Slick]

Originally posted by MicroE:
</font><blockquote><font size="1" face="Verdana, Arial">quote:</font><hr /><font size="2" face="Verdana, Arial">Originally posted by Bandgap:
....5W white has a thin FR4 layer and the central die holder goes down to the aluminium through a clearence hole....

Steve

<font size="2" face="Verdana, Arial">Doug---This is correct. The 5W slug sits directly on a star-shaped sheet of aluminum. There is a thin epoxy layer to hold it in place, as you would expect.---Marc</font><hr /></blockquote><font size="2" face="Verdana, Arial">This is interesting.. I too, have not disassembled a 5W star. If there's a hole through the FR4, then the slug is sitting directly against the heatsink - which I would think might tend to complicate electrically isolating the slug from ground depending on the installation. Time to look over those 5W application briefs again...

 

[Doug S]

Originally posted by Slick:
</font><blockquote><font size="1" face="Verdana, Arial">quote:</font><hr /><font size="2" face="Verdana, Arial">Originally posted by MicroE:
</font><blockquote><font size="1" face="Verdana, Arial">quote:</font><hr /><font size="2" face="Verdana, Arial">Originally posted by Bandgap:
....5W white has a thin FR4 layer and the central die holder goes down to the aluminium through a clearence hole....

Steve

<font size="2" face="Verdana, Arial">Doug---This is correct. The 5W slug sits directly on a star-shaped sheet of aluminum. There is a thin epoxy layer to hold it in place, as you would expect.---Marc</font><hr /></blockquote><font size="2" face="Verdana, Arial">This is interesting.. I too, have not disassembled a 5W star. If there's a hole through the FR4, then the slug is sitting directly against the heatsink - which I would think might tend to complicate electrically isolating the slug from ground depending on the installation. Time to look over those 5W application briefs again...</font><hr /></blockquote><font size="2" face="Verdana, Arial">I wonder if they control the assembly process to ensure that the [non-conducting] epoxy under the slug is thick enough to provide electrical isolation. I did recheck the 5W datasheets to see if that statement from the 1W datasheets is still there about the slug not being electrically neutral. Yep, its still there on the 5W datasheets.