Heat conductive gel (insulator)

Hi.

I'm planning to make a simple power supply. I want to put it in a apropriate box and fill that box with a gel that conduct heat very well and works as an insulator too - the gel is supposed to remove heat.

I've being looking around and so far I found something called Raytech magic gel. The problem is the high price - I can easilly get two apropriate power supplies for that same price.

Is there any alternative gel products - cheaper price per volume of course.

You sink it mineral oil in metal container like high power transformers dielectric constant is =3. Best thermal conductive and electrical insulation but slight capacitance.

clean with alcohol. But why so inefficient?

The problem with mineral oil is that it's fluid. I need some substance that I can fill in a box, and then harden. It may harden into solid or as gel.

Which Raytech Magic Gel did you find suitable, except for price? The one I looked at, MagicGel 300, has a thermal conductivity of 0.2 W/mK (https://uk.farnell.com/raytech/magi...lymer-gel-300-ml/dp/1663039?CMP=GRHB-OCTOPART , then click on Technical Datasheet).

By comparison, Hysol GR750 has a conductivity of 2.1 W/mK.

If 0.2 W/mK is good enough, you could take ordinary epoxy or almost any potting rubber and fill it with fumed silica or alumina and meet or exceed that value. As for cost, the Magic Gel is $20.08/300 mL. That is not terribly expensive compared to similar, two-part RTV silicones.

To put some of the thermal conductivity data into perspective, this chart from Wikipedia might be useful:


John

Material electrical qualities can interfere with thermal when capacitance and resistance degrades performance.
A model of hotspots , Pd, and thermal substrate would assist us greatly. There may be better solutions than what you seek.

Hi.

Maximum cooling and as minimum electrical conductivity is prefered. I'm planning to use it for a very simple power supply, trafo + rectifier.
Input voltage will be in the 230V range, so the absolute most important feature is that the material does not conduct current.

[edit]
After comparing prices, I just think I'll abandon the project. I think that making a simple temperature regulator that controlling air cooling (by fan) is cheaper - after all I have all components needed for that.

Sounds like your parts are overloaded and under-designed/ If you keep it enclosed, it is more likely to overheat from lack of convection cooling and circulation. But ESR in each parts is your only source of excess heat * I^2.

I intended to use this project as learning. In fact I was intended to overload a small trafo and try to find out excactly how much I could load it compared to just air cooling (no fans).
But as said, price is just too high. I'm better off just using a physically bigger trafo.

Or I could just add an old cpu heat sink just for the sake of using it, and have a reason to saw an hole in the box to make the heat sink touch fresh air.
No - I won't do that.

Overtemp in transformers breaks fine secondary wires. A quick way to burn out a transformer is enclose it.... Every Power Electronic designer needs to understand thermal resistance and influence of free air , also linear surface forced air speed effects on thermal resistance.


Used PC and lappy chargers are cheap solutions.