Home made heat sink pads

Silicone is a terrible heat conductor. Filled silicone is slightly less terrible.

One trick is to add a little sand (very little) to epoxy and epoxy the transistors to a heatsink. The sand grains enforce an insulating spacing. This is thermally mediocre, similar to a sil-pad, but useful for low power stuff. (Cataphote makes precision glass balls to do this more scientifically. It's just a screen-graded version of their retroreflector beads.) You can also use thread or monofilament as an epoxy spacer. I suppose sand in grease might work too.

There's a new, relatively expensive, pink kitchen wrap stuff, maybe from Reynolds. It seems pretty tough. A layer of this, with some grease (vaseline?) on each side, wouldn't be bad; just make sure there are no metal spurs or such that might cut through. Thicker plastic, like maybe some household packaging stuff, maybe mylar sheet from a greeting card or some such, would be tougher, but keep in mind that thermal resistance increases directly with thickness.

It would be interesting to compile a list of the thetas of household materials, like yogurt covers or milk carton stock or whatever. Auto supplies (shims, gaskets) would be interesting too.

The best insulator is of course none; float the heat sink and use a dab of grease under the semi.

John

Well, When I did some searches I found many silicone based heat sink pads. While it might not be the best from what I found on the internet it might be good enough for my applications. Here I'm talking about a thickness of less than a mm.

Well, wouldn't a thin piece of hard plastic with many small holes in it work? (say, soemthing from a soda bottle with many 2-3 mm holes drilled in it along with some thermal grease?

Yeah. I tried searching for some possibilities that I could put together quickly and silicone was the only thing that I could find that might work.

There are "Sil-Pads" with Fair+ thermal insulation it has 1 - 1.5 thermal resistance while it says mica is 0.75 to 1.0.

So as far as I can tell you guys are wrong about the silicone. I'm not sure if the sil-pads are "standard" silicone or not but it seems so.

On another site I saw that 1mm of silicone offers about 1kV of protection. Since I just need ~12V I could get away with a pretty thin covering. Just enough to prevent shorting. I doubt the components themselfs would get about

100 C but I wanted to mount them to the chasis just incase they did get to hot.

Thanks, Jon

Out of curiosity: why do you want them home-made? What would be wrong with using some commercially-available insulators? They are not expensive. (Ex. made from some "Boron nitride loaded silicon Elastomer")

Thanks.

-- Andy

But perhaps you don't understand that 1mm is HORRIBLE ?

What's the thermal conductivity of polystyrene though, and in thick layers? Numbers matter.

The see-thru windows commonly used in envelopes are mylar, and I've recently measured some: thicknesses ranged from 0.8 to 2 mils. Metallic wrapping "paper" is super-thin aluminized mylar, which might also do in a pinch. Thetas? Dunno.

You've already received advice to the contrary from experts ... real experts. Dismiss it if you will, but, if so, why bother asking?

Best wishes, James Arthur

1mm is for 1kV+. The data seems to show that its not much worse than mica.

I didn't say they didn't. But the numbers I've seen on the internet suggest something very different than what is suggested here. I've not seen anyone here post any numbers but just say "Your wrong". If the site is wrong then prove it but just don't say stupid shit without backing it up. Theres to many "wanna-be's" here and many cannot be trusted. (Not all, there are some good engineers here but most are not)

I dunno either, which is why I asked. I cannot seem to find any real information on this subject on the net and you guys say what I have found is wrong. I don't really know who to believe.

real experts? How do I know they are real experts? Do you really expect me to take it on faith that everyone in this NG is an expert? I'm sorry. If you want to prove you are an expert than back up what you say with data and references. Don't just say "There wrong and I'm right" because when you do that you look more like a cooke.

If you expect me to take everything you guys say on faith then I'm sorry. It won't happen. I'll listen to what you guys have to say but in the end if it doesn't add up then I'll come to my own conclusions(and if I'm wrong then I'm wrong... atleast I'll find out for myself instead of trying to be a mindless drone like you seem to want me to be).

Anyone can claim to be an expert and many people do it. I'm not saying there are not experts in here or those who have replied are not experts but all I know is that there is some contradictions between what people are telling me and hence someone is wrong. Right now the score is against John and Eeyore because I've found many different sites claiming that silicone works. Its possible that all these sites are ran by the same idiot or that all of them are created by idiots but I would imagine the odds of that happening is pretty slim.

I think the issue you need to get over is that you don't have to take what people anyone says on faith. Its ok to be hard headed or skeptical because else you will be conned. Lots of people try to be experts but few are. Many people try to impress others by acting smart but few are actually smart. Is this happening here? I don't know. I just know that someone's wrong and it would be nice to know if silicone is actually worth using or not so I'll know. Probably the only way it will really get resolved is for me to test it out on my own and see. Ofcourse thats much more work than I'd want to do but...

Anyways, I've heard to many "You can't do this or that" type of BS from newsgroups like this and then when you do it it works. Some people try something once and it doesn't work and they conclude that it can't work and then try and convince others it can't work. Those people are usually easy to spot because they go ape shit when you don't believe them 100%.

Jon

Since you don't want to hear things you don't want to hear, that's what you'll have to do. So don't whine about how much work it's going to be. Silicone pads are OK where a largish thermal resistance is tolerable, as for low power densities and small heatsinks; I use them myself in such applications. But you started this thread by asking for materials you might have around the house, which was actually an interesting question. Have you forgotten the title of this thread?

Look at my web site. Look at the power amp pics I posted to a.b.s.e. Wanna see my water-cooled, 2400 volt, 20 kilowatt pulser?

The critical semiconductors are clamp mounted to a machined-flat, gold-plated, water cooled copper block.

If you newbie into an established ng and ask questions and don't get the answers you expected, and insist that people who are trying to be helpful "prove" that they are right, well, foad. And have a nice day.

John

No, of course you oughtn't take things on faith--you're right to be skeptical. On the other hand, it's clear you don't know enough about the topic to appreciate the sagacity of those whose advice you've already rejected--real engineers, not salesmen, but experienced *users* of these things who've made *real* measurements of real products.

Yes, which is why it's worth listening to an actual expert when you finally encounter him. If 99% of people are fools and you reject 100% of them, you'll be 100% right 99% of the time, but 100% wrong when it really counts. I offered my endorsement; it's that's not enough, check the archives.

That's the problem with polling as a way of resolving conflicting information--reality is not a popularity contest.

Others have already done that hard work--made those tests and measurements over the years--and posted their results here. Those are the results you've already dismissed. If you yourself don't know enough to recognize truth and wisdom when you encounter it, then yes, you'll have to rediscover it for yourself. We've all had that dilemma.

Best wishes, James Arthur

Some additives to epoxy improve the theta, ie slate or aluminium powder. I guess the same might work with grease, and some sort of woven cloth to act as a spacer.

Some quick trials would soon show up what works, sounds messy though.

Well, they're not. Sil-pads suck as heat sink insulators.

What the hell is wrong with mica and heat goop?

Or, like John Larkin suggested, don't use any insulation at all, but just float the heat sink?

Good Luck! Rich

Bitch, bitch, bitch.

So, go someplace where people will tell you what you want to hear.

Good Luck! Rich

hmm... your starting to sound like your getting mad because I don't take everything you say on faith. hmm. I wonder what that means?

I asked if anyone knew of heat sink pads that could be made from common materials. I mentioned silicone and immediately you and Eeyore jumped on that saying that it was crap. I have web sites that disagree with you. Now, instead of proving with reason that the sites are wrong I get insulted into trying to believe you are right. Now maybe most people will roll over and play dead but I won't. Do you really expect me to believe you when you reason through "muscle" instead of logic and observation?

omg, I guess that proves beyond any reasonable doubt that you are an expert in heat sink pads of all kinds of materials?

yeah. I suppose I could just test some stuff and see. Maybe burn up a few regulators and see which ones go out first. Just not really something I'm intersted in doing and I figured someone would know.

"Jon Slaughter"

** Typical silicone pads ( eg from Berquist) have thicknesses of 0.2mm and quote breakdown voltages in the range of 3 to 6 kV. ** Where high device power levels are involved, it is much worse than using thin ( ie 0.002 inch) mica and good thermal compound. For low and moderate device power levels, sil pads work fine.

Neither mica nor silicone are technically good thermal conductors, mica is in fact far poorer than silicone - but mic wins as a thermally conductive insulator because it is possible to use such thin layers of it.

....... Phil

and? Did I say I needed something that was good? You know, what sucks is a relative issue. You seem to have overlooked that. I enver said I needed the best method or even a good one. Just something that "works". Now a piece of shit could work AFAIK. I didn't clarify what I was good or not so all in all you have no clue what I actually want.

um, I have none around and that wasn't my question. If I had the shit don't you think I would have used it?

Because the there is possible contact with the heat sink and the ground. It doesn't matter though. I didn't ask for that and I don't have to explain those details as they have nothing to do with my original question which you seem not to have read.

What I expect you to do is go away and don't come back.

John

HAHA. What a looser. Someone disagree's with you and you get mad. Just what I expect from an "expert".

On Wed, 27 Dec 2006 20:09:59 +0100, Andy Gave us:

Good point. I'm sure that ten packs from digi-key of any given form factor can't be more than a few bucks.

I'm not mad; you haven't anywhere near the skill to make me mad. Hell, you can't even spell. And you're too lazy to make some simple thermal measurements. Try starting over in sci.electronics.basics, and lurk a while before you jump in fighting. This ain't the place for people who know nothing but want to argue about it.

John

On Wed, 27 Dec 2006 22:14:42 GMT, "Jon Slaughter" Gave us:

The question is begged, if one posts a query here about a topic, why would one reject the responses from folks having YEARS of experience in the field? Is not the reason for the poser in this forum to garner responses from experienced individuals?

If you want a REAL good heat sink, get one that has been properly hard anodized. THAT surface is non-conductive, and some thermal cream, and a direct attachment, and you are done!

BTW, the silicone (not silicon) thermal grease is ONLY there to take up tiny, tiny crevices between the two mated surfaces. Most less than a couple mils. The silicone grease is actually a carrier for the powder that has the good thermal properties that get utilized here. The grease itself is not the greatest thermal conductor.

Again, this paste is only for taking up that last 5% of non-contacted surface. The rest of the surfaces should be in "intimate" contact, and that is where most of the conduction work will get done.

Air is a poor thermal conductor. So it the grease, just not as bad. Here is the proof.

Air gaps = hotter part

Grease filled SMALL air gaps maximize transfer.

A large gap between a heat generating part and its sink filled with air is a VERY bad conductor, and the part will run away and fail.

That same part with the gap filled with grease will run away and fail. The grease will not conduct across a large cross section of itself (more than a few mils) as it is not meant to do so.

So. What is learned?

A heat generating part/device needs constant, intimate contact with a thermal sinking device for any period during which it is generating said heat in order to keep operating and not generate dangerous temperatures at any time.

When perfect intimate contact is not possible, the space MUST be minimized as best as possible, and any gaps should be filled with something more conductive than mere air that is not going to dry out, and lose its capacity to fill said gap.

You should go back through and read the responses again. Folks here generally try to help, and you should not immediately discount Usenet posters. Most of us here have been in the industry for a long time, and it would serve you well to heed much of what has been iterated to you. Ignore the DonkeyTard though.