I'm surprised that there is no sign of anyone here having used this stuff, because it looks like it might be perfect.
Panasonic make sheets of it that are readily available, but looking at their website, there isn't quite enough data to tell if it's actually as good as it appears to be. For instance:
Panasonic make sheets of it that are readily available, but looking at their website, there isn't quite enough data to tell if it's actually as good as it appears to be. For instance:
- Would the "A-A" or "A-M" types provide sufficient electrical insulation, or would it be necessary to choose one with a separate insulating layer?
- How much does the insulation affect thermal performance?
- How well does it conform to the typically rough machined surface of a heatsink?
- Is it a good idea to use silicone paste on the non-adhesive side?
- They don't state the through-plane thermal conductivity, so presumably it's much lower than in-plane, but how does that affect the total thermal resistance of a typical interface?
Would the "A-A" or "A-M" types provide sufficient electrical insulation, or would it be necessary to choose one with a separate insulating layer?
They spec it by adhesive thickness, AA gives 2kv, AM gives 1kV
How much does the insulation affect thermal performance?
the insulation defines the performance, thinner is better. I also note they provide no composite numbers.
How well does it conform to the typically rough machined surface of a heatsink?
They don't specify anything like that. Weird.
Is it a good idea to use silicone paste on the non-adhesive side?
i would expect that this stuff should be used in lieu of the silicone.
They don't state the through-plane thermal conductivity, so presumably it's much lower than in-plane, but how does that affect the total thermal resistance of a typical interface?
Through plane won't do too much, I'd still go with the 45 degree model.
However, this stuff bare used instead of silicone grease should be a whopping advantage. A .250 by .250 die, like an IRF 450 hexfet with a 1 mil thick grease under it, would be about 3.2 degrees c per watt rise in the grease, this stuff would basically be zero. Dissipating 10 watts in the 450 would have about 30 degrees C more temp rise than using this sheet. Instant cooler performance.
jn
They spec it by adhesive thickness, AA gives 2kv, AM gives 1kV
How much does the insulation affect thermal performance?
the insulation defines the performance, thinner is better. I also note they provide no composite numbers.
How well does it conform to the typically rough machined surface of a heatsink?
They don't specify anything like that. Weird.
Is it a good idea to use silicone paste on the non-adhesive side?
i would expect that this stuff should be used in lieu of the silicone.
They don't state the through-plane thermal conductivity, so presumably it's much lower than in-plane, but how does that affect the total thermal resistance of a typical interface?
Through plane won't do too much, I'd still go with the 45 degree model.
However, this stuff bare used instead of silicone grease should be a whopping advantage. A .250 by .250 die, like an IRF 450 hexfet with a 1 mil thick grease under it, would be about 3.2 degrees c per watt rise in the grease, this stuff would basically be zero. Dissipating 10 watts in the 450 would have about 30 degrees C more temp rise than using this sheet. Instant cooler performance.
jn
On paper it's enough, yes, but the adhesive is presumably not a tough solid sheet, so I'm concerned that squishing it down onto a surface that isn't perfectly smooth could result in thin parts.
RS stock some, given the price I'm not surprised there has been no mention. 🙁 love to try it but sil-pads or mica and grease will have to do.
the pyrolytic graphite may work as heat spreader - very high in plane conductivity - but thru plane and interface thermal contact issue are likely limitation after price
but it looks like you could make a great cat toy: Magnetically levitating graphite can be moved with laser
but it looks like you could make a great cat toy: Magnetically levitating graphite can be moved with laser
Whole sheets are expensive, but it works out to be <£1 per TO247 sized piece, which is not too extravagant... if it performs well.
- Status
- Not open for further replies.