thermal-epoxy

I plan to upgrade my Huntlight FT-01PJ with a SSC P4 emitter and don't want to buy thermal epoxy because I will only upgrade one flashlight.


I have a few grains of white diamond powder (for polishing surfaces) and I have two-component epoxy.


Diamond is a electrical insulator and its thermal conductivity is five times higher than silver – I wonder if a mixture of diamond powder and epoxy will be a useful thermal-epoxy?


Tom

Maby, in the past I have used 2 part epoxy mixed with the white silicone heat sink commound and it has worked quite well.

I plan on mounting some emitters with a 50/50 mixture of JB Weld and Arctic Alumina Ceramique. The Ceramique by itself has no adhesive properties except for suction. The JB Weld will provide enough adhesive force to hold the emitter in place. I have also read a wide range of reports on the thermal properties of JB Weld. Many claim that it will not work but others have used it to attach heatsinks permanently to CPU's and actually saw a decrease in temp.

I think that mounting a CREE with a star or round MCPCB actually has worse thermal properties than using JB Weld by itself with an emitter.

taschenlampe said:

I plan to upgrade my Huntlight FT-01PJ with a SSC P4 emitter and don't want to buy thermal epoxy because I will only upgrade one flashlight.


I have a few grains of white diamond powder (for polishing surfaces) and I have two-component epoxy.


Diamond is a electrical insulator and its thermal conductivity is five times higher than silver – I wonder if a mixture of diamond powder and epoxy will be a useful thermal-epoxy?


Tom

Click to expand...

Tom, in mixtures such as you are proposing, the resulting thermal conductivity is *almost* entirely determined by the component with the lesser conductivity. Folks that mix heatsink compounds with epoxy mostly get an adhesive with greatly degraded adhesive properties with only slight improvement in thermal conductivity. In all but the most demanding applications, plain epoxy will do fine. I suggest that you heat it slightly, along with the items to be joined, to lower the viscosity enabling a thinner joint. Remember that halving the joint thickness achieves the same result as doubling the compound thermal conductivity.

I think it is great to experiment and thats the tradition of CPF.

However for normal modding, I think it is hard to beat Arctic Alumina two part epoxy. It has proven itself over time and a very thin layer will electrically isolate, get rid of the heat and stick like .....whatever.

Do NOT put too much on. Recently people have been suggesting that as the S-P4 has a + base, you somehow need a thick layer or even two to guarantee electrical isoolation.

Do NOT do that. All you need is a very thin even layer and preferably a light spring to keep even pressure in the 5 minute set up time.

Supposedly making your own thermal epoxy does work well:

http://www.overclockers.com/tips683/

:goodjob:

ace0001a said:

Supposedly making your own thermal epoxy does work well:

http://www.overclockers.com/tips683/

:goodjob:

Click to expand...

The operative word here is ''supposedly''. The experiment design in that link is highly flawed if it's intent is to show that any gain in thermal conductivity resulted from adding thermal compound to the epoxy.

Doug S said:

The operative word here is ''supposedly''. The experiment design in that link is highly flawed if it's intent is to show that any gain in thermal conductivity resulted from adding thermal compound to the epoxy.

Click to expand...

No you can only test one thing at a time and I doubt it even occured to him that Epoxy alone would be enough.


His test was a home made compound aganst a control and it worked OK.


Here is another test but I doubt they used enough samples to prove the results.

[font=Arial, Sans-Serif][size=+1]Epoxy for CPU Cooling[/size][/font]


Another interesting article:
http://www.overclockers.com/tips169/




I have used this it works well. You can also just use the epoxy mix on the edges of the of the sink Epoxy is very srong if applyed corectly even mixed.

Just use straight silicone thermal paste in the middle and use a toothpick to apply a very small amount of JB or similar to the very corners. No worry about adheisive working nor the thermal conductivity. We do it all the time for quick test's.

How about a thin layer of RTV silicone adhesive? I know it will isolate electrically, but will it transfer heat like silicone grease?

On older computers before thermal epoxy was common we used to always put a very small drop of superglue in each corner of the CPU, then thermal grease in the middle, then press the heatsink on there and put a weight on it for 5 minutes or so to make sure the superglue was fully bonded. It worked great!

Thank you guys for your tips and opinions - I will keep them in mind …



Tom

At work we use two part epoxy mixed with zinc powder 50/50 to attched themocouples to product parts during temperature rise tests. So I used the same when fixing a 5W emitter to an o sink for my spaceneedle clone, works fine.
Although I'm not sure how/if it would work with a SSC P4 coductive bottom?