Thermal insulators "Phase Change Material"

These are not usually considered to be insulators. We tried phase change stuff and hated it.

Hard anodize on the heat sink, with grease, is about the best you'll do if you need insulation.

More extreme: use a heat spreader block and insulate that from the heat sink.

Even better, don't insulate the transistors from the heat sink.

Den torsdag den 25. september 2014 23.17.53 UTC+2 skrev Harry D:

afaict tell from the datasheet at maximum compression is is 0.013 C-inch^2/W

a TO247 is less that half an inch^2

-Lasse

Perhaps a stupid thought, but are you sure they are intended for high voltage applications? I find it suspicious that they don't specify a breakdown voltage in the datasheet. A high volume resistivity doesn't imply a high breakdown voltage. The application examples are all low voltage, too.

Klaus

I agree, it's not an insulator. Once it changes phase, the insulator properties probably go to hell.

Cheers

They are classified as "Thermal-Interface-Materials", not an insulator.

Cheers

Den torsdag den 25. september 2014 23.17.53 UTC+2 skrev Harry D:

afaict tell from the datasheet at maximum compression is is 0.013 C-inch^2/W

a TO247 is less that half an inch^2

-Lasse So if 0.5 the area is that not 2x the thermal resistance?

Harry

These are not usually considered to be insulators. We tried phase change stuff and hated it.

Hard anodize on the heat sink, with grease, is about the best you'll do if you need insulation.

More extreme: use a heat spreader block and insulate that from the heat sink.

Even better, don't insulate the transistors from the heat sink.

Den fredag den 26. september 2014 01.53.22 UTC+2 skrev Harry D:

yes so 0.026 you said you need 0.016

-Lasse

No magic, Harry Potter. Bolt the transistor to a copper block, and insulate the block from the heat sink. Theta goes down because the insulator area increases. You get a bonus reduction in theta by reducing the spreading thermal resistance of the heat sink itself.

Another possibility is to make the spreader block out of aluminum, bare on top and anodized on the bottom, for the insulation.

All machined flat and greased, of course. The main heat sink needs to be flat, too.

Here's an aluminum heatsink with copper heat spreaders. No insulators... the heat sink itself is mounted on delrin blocks.

You mentioned 150 VDC. I wouldn't recommend anodize above 200 volts.

Oh, a good high-voltage insulator is greased aluminum nitride. Might need a heat spreader.

Dow 340 grease will squeeze down to under 100 micro-inches (my measurement resolution) with moderate pressure. The phase-change gorp oozes out a bit around the edges but stays mills thick, so has a lot more theta than grease. Neither insulates.

If you replace a transistor, the gorp is messier than the grease.

Kapton is a great insulator, but theta is huge.