Maker Pro
Maker Pro

Regular epoxy adhesives for heat sink?

primuspaul

Feb 7, 2018
91
Joined
Feb 7, 2018
Messages
91
I recall reading that regular epoxies transfer heat about as well as the more specialized heat sink / CPU adhesives and work sufficiently well provided that the LED/chip is as close to the heat sink as possible. Is this accurate?

If so, that provides a much cheaper solution (and one which doesn't suffer from a short shelf life thanks to being a binary compound) to attaching heat sinks to chips/LEDs.
 

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
Epoxy will glue the devise to the heat sink and you will not be able to dismantle it, should it go wrong.
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
Somewhere I pointed to an article where heatsink compound was compared to various alternatives.

Some of these alternatives (if I remember correctly) were toothpaste, butter, and chocolate.

In a not very surprising result, they found that almost anything worked. The reason is that the primary determinates of efficiency are:
  1. Being better at transferring heat than air
  2. Being able to form a thin layer which replaced air.
An adhesive used to mount something to a heatsink must also be able to do the above. To be effective as an adhesive, the second property is pretty much a requirement, and since air is an excellent insulator, almost anything meets the first condition above.

As in the experiment I referred to, where commercial products design specifically for the purpose were found to be better, I expect you would find the same. However, if you design conservatively, the difference may be able to be tolerated.

In the experiment I mentioned, they also commented on the fact that commercial products are designed to have long term stability in this use. They did not test, but we're reasonably confident that things like butter and toothpaste would degrade pretty quickly. In your case, adhesives are typically designed for long term use and my only concern would be the effect of maintaining long term elevated temperatures. But it may be possible to get this information from the manufacturer assuming you don't buy a no name product.
 

primuspaul

Feb 7, 2018
91
Joined
Feb 7, 2018
Messages
91
Epoxy will glue the devise to the heat sink and you will not be able to dismantle it, should it go wrong.
Yes, obviously it is not a good idea to use this method for Intel CPUs, but then it's not necessary either since those are bolted/clipped down and thermal paste is sufficient. I am referring more towards $0.30 LED chips. Those only have to be removed when they go bad and the heat sink surface can simply be filed down a bit to remove the epoxy (once the bad LED chip is ripped off).
 

primuspaul

Feb 7, 2018
91
Joined
Feb 7, 2018
Messages
91
Somewhere I pointed to an article where heatsink compound was compared to various alternatives.

Some of these alternatives (if I remember correctly) were toothpaste, butter, and chocolate.

In a not very surprising result, they found that almost anything worked. The reason is that the primary determinates of efficiency are:
  1. Being better at transferring heat than air
  2. Being able to form a thin layer which replaced air.
An adhesive used to mount something to a heatsink must also be able to do the above. To be effective as an adhesive, the second property is pretty much a requirement, and since air is an excellent insulator, almost anything meets the first condition above.

As in the experiment I referred to, where commercial products design specifically for the purpose were found to be better, I expect you would find the same. However, if you design conservatively, the difference may be able to be tolerated.

In the experiment I mentioned, they also commented on the fact that commercial products are designed to have long term stability in this use. They did not test, but we're reasonably confident that things like butter and toothpaste would degrade pretty quickly. In your case, adhesives are typically designed for long term use and my only concern would be the effect of maintaining long term elevated temperatures. But it may be possible to get this information from the manufacturer assuming you don't buy a no name product.
I had golf club epoxies in mind, but yes, I also considered using no-name Chinese epoxies. I am optimistic since epoxies are generally considered to be the strongest of adhesives (some being used for golf clubs) and since holding an LED chip in place requires minimal strength (even thermal paste is enough).
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Jan 21, 2010
25,510
Joined
Jan 21, 2010
Messages
25,510
I had golf club epoxies in mind

Golf club epoxies are probably not the sort of epoxy intended to be exposed to continuous high temperatures.

I would find a brand name adhesive that has an available datasheet.

For example, for loctite products, see here.

As an example, loctite 4090 has this datasheet. Compare that with Loctite 7234 (here)

The former example shows it will fail at about 180C-200C (which is way beyond the maximum temperature you should ever hope to reach. The latter doesn't give graphs, but states it is fine up to 220C, however beware that it recommends the adhesive layer is 0.5mm in thickness.

loctite 680 would probably be a good alternative (datasheet). It is stronger when applied more thinly, and has excellent aging properties at temperatures you are unlikely to exceed.
 
Top