All you wanted to know about Radio Shack and other silicone thermal compound

[Jerboy]

Yo people. I was reading a thread about Radio Shack thermal compound on Off Topic and we were debating whether it is reasonably good or not. I've done some investigation on Radio Shack compound and came up with some results and just wanted to share the outcome with you.


Radio Shack sells heat sink thermal compound in small tube for $1.99 or so and its product number is 276-1372. As you might have guessed, it's OEM.


The manufacture is called GC Thorsen and the chemical composition is 70% silicone oil(CAS# 63148-62-9) and 30% zinc oxide. Zinc oxide is the stuff that actually conducts heat. Zinc oxide by itself is heavy white powder. Silicone oil is stirred with zinc oxide to bind the powder together and preventing the release of zinc oxide in dust form.

Radio Shack thermal compound Material Safety Data Sheet

High quality silicone thermal compound from a reputable brand such as Techspray has this composition:

Zinc oxide content >70%
remaining was not identified

Techspray MSDS


Since ZnO is more expensive than silicone per volume, they will use smaller percentage of zinc oxide to cut cost. Anyone whose used the Radio Shack compound will know how liquidy it is and I think I pretty much explained why it is that way.

 

[Richardito]

How do you know the RS compound is 30% ZnO? The price of the ZnO depends on the grade and mean particle size. I know of a ZnO grade that costs nearly $20 a pound (not cheap), while coating industry standard is dirt cheap at less than 1/2 of that. It may or may not be more inexpensive than the silicone fluid, it all depends. I don't agree when you mention that the only reason for the fluid is to keep the ZnO in one place. The silicone fluid is also used to transfer the heat, since air is a very good insulator and any liquid is a better heat transfer medium. Even if you place ZnO on top of your CPU and placed your HS on top of it, the thermal transfer will be much worse than with the liquid thermal compound. Take a look at this Table. I do dispersions everyday...

 

[Jerboy]

<< How do you know the RS compound is 30% ZnO? >>



I have not done a quantitive test on it myself. It is just what they declared on MSDS



<< The price of the ZnO depends on the grade and mean particle size. I know of a ZnO grade that costs nearly $20 a pound (not cheap), while coating industry standard is dirt cheap at less than 1/2 of that. >>



I'd be surprised if Radio Shack used premium quality stuff.




<< It may or may not be more inexpensive than the silicone fluid, it all depends. I don't agree when you mention that the only reason for the fluid is to keep the ZnO in one place. >>



The silicone is pretty much there to keep things together. Pure ZnO won't stay in place. It's a good thermal conductor, but it won't make good thermal grease by itself.



<< The silicone fluid is also used to transfer the heat, since air is a very good insulator and any liquid is a better heat transfer medium. Even if you place ZnO on top of your CPU and placed your HS on top of it, the thermal transfer will be much worse than with the liquid thermal compound. Take a look at this Table. I do dispersions everyday... >>




What difference does it make if you use things like soap, KY Jelly, sugar syrup or motor oil in a short run? I woudn't expect a whole lot.

 

[Richardito]

Originally posted by: Jerboy

How do you know the RS compound is 30% ZnO?

Yep, it's 30%.

I have not done a quantitive test on it myself. It is just what they declared on MSDS

The price of the ZnO depends on the grade and mean particle size. I know of a ZnO grade that costs nearly $20 a pound (not cheap), while coating industry standard is dirt cheap at less than 1/2 of that.

I'd be surprised if Radio Shack used premium quality stuff.

The most costly one doesn't cost more because it is 'premium quality', it costs more because of it's particle size and application (cosmetics and sun tan lotions). ZnO is very cheap stuff. I'll get a price quote on Monday and verify the lowest-cost formulation with that information.

It may or may not be more inexpensive than the silicone fluid, it all depends. I don't agree when you mention that the only reason for the fluid is to keep the ZnO in one place.

The silicone is pretty much there to keep things together. Pure ZnO won't stay in place. It's a good thermal conductor, but it won't make good thermal grease by itself.

No, the other reason it is used is because it is an insulator. Also depending on the grade that is chosen (different molecular weights) other things such as viscosity can be changed. Have you ever made thermal compound formulations in a lab? Do you know the basic dispersion dynamics?

The silicone fluid is also used to transfer the heat, since air is a very good insulator and any liquid is a better heat transfer medium. Even if you place ZnO on top of your CPU and placed your HS on top of it, the thermal transfer will be much worse than with the liquid thermal compound. Take a look at this Table. I do dispersions everyday...

What difference does it make if you use things like soap, KY Jelly, sugar syrup or motor oil in a short run? I woudn't expect a whole lot.

Again, silicone fluid is safer because it doesn't conducts electricity. If it doesn't matter, then why does the mayority of companies use it to manufacture thermal compounds? It is not cheap either...

 

[Richardito]

I just got the quotes: Dow Corning 510 silicone fluid ($75 per pound), filler grade ZnO ($0.80 per pound). As you can see the ZnO is much cheaper that the silicone fluid. So how is that using less ZnO makes a lower cost product for RS? Actually the more ZnO a thermal paste has the cheaper it is (raw material cost-wise).

 

[lamga]

The most costly one doesn't cost more because it is 'premium quality', it costs more because of it's particle size and application (cosmetics and sun tan lotions

Hmm, I wonder if anyone should do a thermal compound test but with Coppertone? SPF30 vs SPF50? Like at Dan's Data, with AS3, toothpaste, PB and Vegimite (toothpaste won, but also dried out in a few hours)!

 

[Mikewarrior2]

its just too bad DansData's test rig isn't more analogous in size to current CPU die's.



Mike