heat dissipation method

[Project16]

Hello all!

What is the best method for electrical isolation and heat dissipation in a final stage heating much?
I personally did not know that the traditional mica more silicone grease but I found other methods.

http://www.bomberg.com/Log/Files/Keratherm_eng.pdf

thank you.

Sorry for my english!

[gmphadte]

You might consider using copper plate fixed to your heatsink, to improve heat dissipation.

Gajanan Phadte

[JMFahey]

From the Keratherm graph on Page 7, mica + grease is better than even the best of their materials , Keratherm Red , so ... I'll keep using mica + grease in the foreseeable future.

[N-Brock]

Originally Posted by JMFahey

From the Keratherm graph on Page 7, mica + grease is better than even the best of their materials , Keratherm Red , so ... I'll keep using mica + grease in the foreseeable future.

I take that graph to mean that just thermal grease was more thermally conductive.

Nothing is mentioned about the mica as far as I can see.

Best,
N-Brock

[JMFahey]

You are right, they don't *specifically* mention mica, ... and why would they?
Who wants to push "competition"?

But mica is not a "mystery" material, being the standard for many years now against which all others are compared to.
And its value can be "read" (or better said, "interpolated") from that graph easy enough.

A typical Rth value for Mica is around twice what would be achieved by grease alone, as in the old rating shown over and over in datasheets, since when Germanium Transistors were the Industry standard.

As in: "Thermal resistance case to heat sink, 0.5ºC/W with grease alone ; add 1ºC extra if mica (+ grease, of course) is used"
So I consider (for quick comparison, of course) mica +grease=3X grease alone.

And even Keratherm red does nor seem to beat that.

Can't be more precise, because that is not a proper Rth graph but a 3D bar/columns graph, so you can't exactly match column height to the scale on the left.

They do show a product which seems to beat grease + mica, which is some kind of wax (PCM).
Probably it's useful in some very special situations, but I would be very nervous using a thermal/insulating material that melts between 48 and 60ºC

[DF96]

Most modern replacement for mica+grease are aimed at quicker or more convenient assembly in a factory, not better performance. It may be that they work better when badly used, while the mica method needs to be done properly: flat heatsink (no burrs etc.), very thin layer of grease, no air pockets.

[JMFahey]

Fully agree.

[frank1]

I will also stick with mica.
You should realise though that "standard" mica washers are much thicker and have a much higher voltage breakdown than is normally needed.
You can carefully de-laminate mica washers to reduce their thickness considerably and thus reduce Rth.

Frank

[DUG]

Originally Posted by DF96

Most modern replacement for mica+grease are aimed at quicker or more convenient assembly in a factory, not better performance. It may be that they work better when badly used, while the mica method needs to be done properly: flat heatsink (no burrs etc.), very thin layer of grease, no air pockets.

Fully agree #2

[DUG]

Originally Posted by frank1

...You can carefully de-laminate mica washers to reduce their thickness considerably and thus reduce Rth.

Frank

Heavy emphasis on the carefully.

If you don't get even layers then you will stress the package during mounting.

But a good idea for lower thermal resistance.