open can LD double side cooling?

[CHP]

If you look at the thermal structure of the LD chip, it only gets cooled from one side. Is there something (electrically insulating, thermally conductive epoxy??) that can be placed on the top of the LD chip to help get heat out from the top side. It seems like any thermal parallel path would help get the heat out from the source.

 

[flogged]

Try arctic silver epoxy:
http://www.arcticsilver.com/arctic_silver_thermal_adhesive.htm

Supposedly CompUSA and RatShack sometimes carry it. For my next DIY red I'll be using this glue to bond the diode to the case housing to ensure it stays centered in the optical path.

 

[Gazoo]

Try arctic silver epoxy:
http://www.arcticsilver.com/arctic_silver_thermal_adhesive.htm

Supposedly CompUSA and RatShack sometimes carry it. For my next DIY red I'll be using this glue to bond the diode to the case housing to ensure it stays centered in the optical path.

I do not understand why you would want to do this. The diode presses in tightly so why do you need anything else to hold it in place? Also the residue from the glue as it is drying could mess up the glass that covers the diode...with an open can it would be worse.

 

[flogged]

[quote author=flogged link=1199137573/0#1 date=1199138647]Try arctic silver epoxy:
http://www.arcticsilver.com/arctic_silver_thermal_adhesive.htm

Supposedly CompUSA and RatShack sometimes carry it. For my next DIY red I'll be using this glue to bond the diode to the case housing to ensure it stays centered in the optical path.

I do not understand why you would want to do this. The diode presses in tightly so why do you need anything else to hold it in place? Also the residue from the glue as it is drying could mess up the glass that covers the diode...with an open can it would be worse. [/quote]

I'm not doing it, the OP is. I'm using a Meredith LDC-1 lens assembly and I've found the press fit is not tight enough to prevent the diode from squirming while focusing the lens. I'd be gluing the back end of the 5->9 mm adapter to the back of the case and this glue is ideal because it would not block the thermal path. This is the primary way heat exits the diode to the case. I'm not talking about gluing anything to the front of the diode.

 

[Phazonx]

[quote author=flogged link=1199137573/0#1 date=1199138647]Also the residue from the glue as it is drying could mess up the glass that covers the diode...with an open can it would be worse.

Yeah, that makes my skin crawl. Just thinking about dropping that stuff right on an open can...
Its an expensive mistake.

 

[CHP]

My initial question dealt with the possibility of double side cooling the LD chip (similar in concept to the double side cooling of a hockey puck SCR package). I was looking for something that doesn't give off fumes (that could possibly coat the output coupler) and doesn't conduct electricity.

Some of the lower power LD can have a junction-to-case thermal impedance of 250 C/W. So if the case is at 70 C the junction can be 95 C. If you look at the thermal network from junction-to-ambient the net thermal impedance is dominated by the junction-to-case thermal impedance. The goal of my question was to attack the most dominate thermal impedance in the thermal network to minimize the LD junction temperature and permit higher power operation (assuming no optical damage).

So, even though there is probably very little thermal stratification perpendicular to the junction, providing a parallel thermal path from the exposed side of the LD chip could offer a significant reduction in the junction-to-case thermal impedance. A 20% reduction in the junction-to-case thermal impedance is worth more than having an ideal heatsink connection to room temperature on the aixiz case.

 

[wooooooolazer]

My initial question dealt with the possibility of double side cooling the LD chip (similar in concept to the double side cooling of a hockey puck SCR package).  I was looking for something that doesn't give off fumes (that could possibly coat the output coupler) and doesn't conduct electricity.  

Some of the lower power LD can have a junction-to-case thermal impedance of 250 C/W.  So if the case is at 70 C the junction can be 95 C.   If you look at the thermal network from junction-to-ambient the net thermal impedance is dominated by the junction-to-case thermal impedance.  The goal of my question was to attack the most dominate thermal impedance in the thermal network to minimize the LD junction temperature and permit higher power operation (assuming no optical damage).  

So, even though there is probably very little thermal stratification perpendicular to the junction, providing a parallel thermal path from the exposed side of the LD chip could offer a significant reduction in the junction-to-case thermal impedance.  A 20% reduction in the junction-to-case thermal impedance is worth more than having an ideal heatsink connection to room temperature on the aixiz case.

heh, my brain hurts...

...lazer...