Thermal Grease/Adhesive (?)

Years ago ( Jesus, about 40 actually ... ) a buddy of mine in chemistry class, got asked by the teacher, what the chemical symbol ' Sn ' was. Quick as a flash, he said " Silicon, sir ! "

After a suitable pause, the teacher said " Aye, Smith, you really are a silicon, aren't you ... ? "

His name wasn't actually Smith ...

Arfa |:-o

I have a Leach audio amplifier built in the 1970s, and a few years ago its Radio Shack or GC silicone grease hadn't dried out. So I wouldn't worry about this, and I don't see why Arctic Silver would be more resistant to drying out (it may use ester instead of silicone grease).

It depends on how hot the semicondictor runs. It's possible the effectiveness of thermal compound can deteriorate as it changes state.

It is based on polysynthetic oils. The manufacturers apparently claim it won't dry out, and my experience with it certainly backs their claims up- I have never seen dried Arctic Silver.

In any case, I am not advising anyone to rush out and buy Arctic Silver for electronic work. I merely meant that for the OP's purposes, any form of thermal compound he could get his hands on would do the job. Plain old white, silver, whatever. Many people build their own PCs these days, so I thought perhaps he may have a tube of CPU compound lying around, or have a nearby computer shop he could get it from.

Dave

I have seen the white stuff dry out plenty of times. Its sure easy to spot and feel. If you can't smear it then it must be removed. If its still good, I would just even out the stuff and reapply components.

greg

I agree in theory. But in practice, thermal grease is so cheap compared to what it prevents that it is foolish economy IMHO.

Agreed.

If there is some small particle in the thermal compound, it would separate the component from the heatsink and could dramatically decrease cooling.

The best plan is to carefully clean both surfaces and use new, clean compound.

I was building amps from STK modules. I studied the mating surfaces. Ideally metal to metal contact is best, even the anodized surfaces should be milled down. So I was taking the STK module and sanding it down flat. It really wasn't very flat. I had to start out with a flat surface for the sandpaper! After much work, I got it to the point I felt was OK. Even a flat etched surface is worse than a mirrored flat surface. Very very little grease is needed if everything is smooth and flat.

greg

I noticed the CPU temperature on my main computer going a bit high a few months ago. I checked, and sure enough the heat sink grease had dried out. Replaced it and all was well.

Disagree. Where it's needed, it's critical; it's cheap enough that replacing it in critical places is well worth while.

They don't suffer from thermal problem when mounted conventionally, with a dab of heatsink compound, so why bother ?!!!

All a bit OTT consuidering the STK modules don't sound that great in the first place.

geoff

Sounds good enough for the common man. Running the 200 watt modules, now extinct, but rugged as hell. I ran at reduced voltage for 200 watts at 4 ohms. I was actually driving nearer 3 ohms at one time.

greg

i suppose the fluid helps to also fill in gaps having to fit into the compound mix of dry materials. Having dried out may not always be a problem. There should be very little actual compound that sits between flat surfaces. Its almost totally trapped in. if the surfaces are poor, and the strength of the compression is poor, I think it would dry out quicker. Probably a good choice for gasket type pads.

greg

"Richard Crowley" wrote in news: snipped-for-privacy@corp.supernews.com:

What dries out (white stuff) is not the thermal conducting material. It should be reapplied solely for best thermal contact. And the thinnest,most even layer is most optimal.

Makes me feel good that Usenet is available even to the insane.