Re: Inretech Trilight3
> And yes, it did feel kinda hot in my hand, but still well below the 212 degree F the emitters can reportedly handle.
Actually, you don't know what the junction temperature is of the semiconductor material, which is the important place to measure. It is probably very close to the 135C spec, possibly even over the spec.
Here is a some info on junction temps vs externally measured temps.
The thermal resistance between the semiconductor (LED itself) and the outside world must be taken into account if you are not measuring temperature at the semiconductor.
If you measure the temperature of the flashlight head, then what you are measuring is the temperature drop between the case and the air:
air temp | res | case temp | res | module temp | res | LED
The above is meant to illustrate that there are various thermal resistances between the elements. There are of course many more thermal resistors in there depending on the design of the system. The Luxeon LED is mounted on a slab of aluminum, and there is thus a built-in thermal resistor between the LED and this aluminum, before you even add anything else to the system. For one of the Luxeon die mounted on an aluminum heat spreader, the thermal resistance is about 11 degrees C/watt. So, if you have a raw 5W Luxeon Star, and you measure an aluminum slug temp of say 30C, then the *actual* die temp might be 30C + 11C/W * 5W = 85C.
If you have a Luxeon mounted to some kind of assembly which is then mounted inside of a Mag head, the thermal conductivity probably isn't really great and thus the thermal resistance is higher than you would like or think. You could easily be adding a thermal resistance of say 20-30C/W in there, in which case when used with a single 5W LED the temp at the LED substrate would increase by 150C compared to the barrel temp.
That is one reason my design mounts the Luxeon straight onto a heatsink, and the heatsink acts as part of the housing. I wanted to get the lowest thermal resistance I could out of the system. The heat sink adds something like 8C/W on top of the 10C/W or so that the aluminum slug adds, for a total of 20C/W. When I run it at 5W, the heatsink temperature gets up to about 130F after 18 minutes (measured with an infrared thermometer).
In your case, you measured 14.1W going into the head, and 49C case temp. We know that there is at least an 11 C/W thermal resistor in there. I assume that there are 3 in parallel (3 LEDs?). Assuming they aren't on the aluminum heat spreaders but are emitters, then the thermal resistance drops to 8C/W. My heatsink is only around 8 C/W, and I suspect that your flashlight has a higher thermal resistance than that, however lets just say it's also 8C/W. 49C + 14.1 * (8/3 + 8) = 199.4C. Hmm, I think you are over the 135C limit.
For more info see
http://www.theledlight.com/pdf/Luxeon/Technical5W.pdf
Scott