Thermal Test Results - PCB vs Aluminum Heatsink

Hi All,

I've not posted much lately; I've been on vacation and just enjoying the nice weather for once. /ubbthreads/images/graemlins/cool.gif

However, I did manage to finally get the thermal test posted which compares the original Illuminator Pill (PCB heatsink) against a MiniPro (aluminum heatsink).

Here's a snip from the testing:


At 700mA the PCB unit only produces about 83% of initial brightness after 10 minutes. The Heatsink unit produces about 93% of initial brightness; 10% brighter than the PCB test unit.

Many more charts with data from 100mA to 1200mA in the complete test charts located on the Thermal Test web page

Enjoy,

Lambda

Re: Thermal Test Results - PCB vs Aluminum Heatsin

That's the reason why I built Copper Heat Sink...

I wonder if they will fit on your Pills.... If it works... that will be a Great thing /ubbthreads/images/graemlins/smile.gif

Great work, Lambda! Those runs must have taken quite a bit of time to perform. Although very applicable to the LuxIII, these results will become even more important for the higher current of the upcoming K2. Thanks for sharing your work. /ubbthreads/images/graemlins/happy14.gif

Paul

chimo - that's an understatement; it took forever with having to wait between each test for everything to cool back to room temp; it took several days...

ViReN - Great heatsinks!

It looks from the photo they would need a little more boring on the back side to make room for the MiniPro converter though. /ubbthreads/images/graemlins/frown.gif

But that would make an interesting comparison, copper heatsink against aluminum heatsink; I wonder how much better the copper might be? /ubbthreads/images/graemlins/thinking.gif

Awesome - thanks for taking the time to collect all of this data /ubbthreads/images/graemlins/happy14.gif

Will

Re: Thermal Test Results - PCB vs Aluminum Heatsin

So is there a good conductive thermal paste that would help get the heat from the heatsink to the mag body? It seems that this is one area where improvements may be found.

Re: Thermal Test Results - PCB vs Aluminum Heatsin

Thanks Lambda!
Always liked your heatsinked pills, it makes sense to absorb the heat and transfer it to the minimag. I can vouch for the effectiveness of PCBs at low drive currents...your Lambda MicroIlluminator runs nice and cool and has done so for the last two years. (R3J Luxeon running at 155mA)
Now that we have the truth, when will the drop-in versions of your heat sinked pills start rolling off the assembly line? Make mine a Luxeon K2 SWAH running 700mA /ubbthreads/images/graemlins/takeit.gif /ubbthreads/images/graemlins/happy14.gif

Re: Thermal Test Results - PCB vs Aluminum Heatsin

Lambda,

Great to see your post and awesome data too! Very useful in future designs of high powered LED sammie or the like. /ubbthreads/images/graemlins/thumbsup.gif

Re: Thermal Test Results - PCB vs Aluminum Heatsin

Very very valuable data! Cheers!

Re: Thermal Test Results - PCB vs Aluminum Heatsin

Lambda,

Thank you for providing your collected data to the community here. Nicely thought out and well done.

Curious, you didn't mention how thick the copper clading was on the PCB. Most these days are 1/2 oz. copper, a person can purchase up to 4 oz. copper to reduce the thermal resistance in the board by a factor of 8x.

Additionally, besides the the thermal resistance going across the 0.000707" (4 oz is 0.0057" as I recall) layer of copper on the board, there is also the contact area to the walls of the flashlight, another point of thermal resistance. It appears Viren's slug has a substantial wall which helps address this second issue.

As to the copper vs. aluminum questions, here are my thoughts on it from January 2004:

So, for volume:
- Copper will soak up 142% more heat than Aluminum
- it will have half the thermal resistance at
the same time.
- average 40% better electrical conductivity per size (or 40% less electrical resistance.)


Keep in mind, most numbers a person finds are for pure aluminum, and not the alloys everyone typically utilizes, which are even worse, you add 20% more thermal resistance, with some common aluminum alloys having 1/3 the thermal conductivity of copper.

Even LumiLEDs switched over from aluminum slugs to copper slugs on their emitters a couple of years ago.

Additional information is found in the old thread here:
http://www.candlepowerforums.com/ubbthreads/showflat.php?Cat=&Board=UBB14&Number=448357

Re: Thermal Test Results - PCB vs Aluminum Heatsin

[ QUOTE ]

........
...there is also the contact area to the walls of the flashlight, another point of thermal resistance. It appears Viren's slug has a substantial wall which helps address this second issue.
........


[/ QUOTE ]

That is one of the reasons why I kept the Walls Thick /ubbthreads/images/graemlins/smile.gif, Just a Disc wont suffice more contact to the sides will definately help.

Re: Thermal Test Results - PCB vs Aluminum Heatsin

Pure copper can be a PITA to machine. It's very gummy and can require a lot of deburring. There's an alloy called Telurium copper (CDA 145) which is free machining almost like leaded brasses, and doesn't give up too much in thermal conductivity. McMaster carries it.

Microcans
Minicans

Larry

Re: Thermal Test Results - PCB vs Aluminum Heatsin

tvodrd: Wow.... Great Cu Heat Sink & Water Cooling System.

Re: Thermal Test Results - PCB vs Aluminum Heatsin

/ubbthreads/images/graemlins/blush.gif Thanks! I sent you a PM with a link to some more pics of the box.

Larry

Re: Thermal Test Results - PCB vs Aluminum Heatsin

wow larry, what is that beast for!?

*edit* what? link? more pictures?

Re: Thermal Test Results - PCB vs Aluminum Heatsin

/ubbthreads/images/graemlins/nana.gif

Larry

tvodrd - What are you putting those Microcans and Minicans into? I'm looking for heat sink/emitter ideas to change my 3D mags into LED's.

Thanks.

lamda, nice test data!