Barbarin
Flashlight Enthusiast
This is part of a post I have written on another thread, but I think you could find it interesting, and I will find to share your ideas.
"Thermal management:
I have been always very concerned about the thermal management of the heat generated by the diodes as it can result on output loose if not designed properly. Most of the people won't use a flashlight more than 30 hours, and at that point won't be noticeable, but from just 100 hours you can detect even a 30% lumen if heat is not managed properly. The cooler the diode runs, the better, and the initial moment, when you switch it on is a critical moment on thermal stress. Heat is generated instantaneously on the die so you have to take it out as soon as possible.
That is why I have designed a intercooler sandwich based on a thick pure copper disc, as copper is near double than pure aluminium, and double than alloys:
Ag (silver) 420 W/(m K)
Cu 400 W/(m K)
Al 240 W/(m K)
(Al, Si, Mg) Alloys 200 W/(m K) average.
Of course a very important point is the joint between that copper disc and the diodes, but even using the best thermal epoxi (arctic) the conductivity of this compounds is 30 times worse than copper itself, arround 8 W/(m K), so we need to get this epoxi layer as thin as possible. In my case I have designed a simple tool wich press the diode against the disc with no less than 15Kg/cm2 during the curing process to make that heat barrier as thin as possible.
Once the heat is on the copper disc it needs to be redirected to the flashlight body itself which in my opinion should be always made of aluminum alloys on diode based flashlights as thermal conductivity of other materials is quite poor when compared:
Steel 45 W/(m K)
Brass 85 W/(m K)
Titanium 22 W/(m K)
To get that trasnference as good as possible teh best way to do it to have the larger area, polished, with a interface material (micronized copper grease in my case) and to press it firmly to minimize the heat barrier.
So once we have that heat on the head of the flashlight a good design to get it redirected to the hole body, and finally to surrounding air or water, is the key the have our porr diodes running as cool as possible. In my case I have done a large threaded area between head and flashlight body, with no o-rings at all, but thin compound, and thsi area is 125cm2.
As a result the hole flashlight becomes warm quickly, and the temperature of the tailcap is similar to the head"
Javier Lopez
"Thermal management:
I have been always very concerned about the thermal management of the heat generated by the diodes as it can result on output loose if not designed properly. Most of the people won't use a flashlight more than 30 hours, and at that point won't be noticeable, but from just 100 hours you can detect even a 30% lumen if heat is not managed properly. The cooler the diode runs, the better, and the initial moment, when you switch it on is a critical moment on thermal stress. Heat is generated instantaneously on the die so you have to take it out as soon as possible.
That is why I have designed a intercooler sandwich based on a thick pure copper disc, as copper is near double than pure aluminium, and double than alloys:
Ag (silver) 420 W/(m K)
Cu 400 W/(m K)
Al 240 W/(m K)
(Al, Si, Mg) Alloys 200 W/(m K) average.
Of course a very important point is the joint between that copper disc and the diodes, but even using the best thermal epoxi (arctic) the conductivity of this compounds is 30 times worse than copper itself, arround 8 W/(m K), so we need to get this epoxi layer as thin as possible. In my case I have designed a simple tool wich press the diode against the disc with no less than 15Kg/cm2 during the curing process to make that heat barrier as thin as possible.
Once the heat is on the copper disc it needs to be redirected to the flashlight body itself which in my opinion should be always made of aluminum alloys on diode based flashlights as thermal conductivity of other materials is quite poor when compared:
Steel 45 W/(m K)
Brass 85 W/(m K)
Titanium 22 W/(m K)
To get that trasnference as good as possible teh best way to do it to have the larger area, polished, with a interface material (micronized copper grease in my case) and to press it firmly to minimize the heat barrier.
So once we have that heat on the head of the flashlight a good design to get it redirected to the hole body, and finally to surrounding air or water, is the key the have our porr diodes running as cool as possible. In my case I have done a large threaded area between head and flashlight body, with no o-rings at all, but thin compound, and thsi area is 125cm2.
As a result the hole flashlight becomes warm quickly, and the temperature of the tailcap is similar to the head"
Javier Lopez
