flexible/ low modulus highish thermal conductivity potting compound

[Siwastaja]

air is ~ 0.02 ?

Thermal conductivity number of air is completely meaningless, as any temperature difference drives convection. That's why it's very wrong to say that air is "thermal insulator", and assume that any potting compound would be automagically better. For example, laptop power supplies do great job of dissipating a lot of power within small frame, and they do not do that by potting, they use heat spreaders (metal or pyrolytic graphite, which is 3-5 times better than pure copper, but only works as thin sheets) to couple that heat over larger area of internal air and external plastic case, both pretty poor thermal conductors, but given enough cross-sectional area, it works out, therefore the initial heat spreading is the key to success.

[tszaboo]

air is ~ 0.02 ?

Thermal conductivity number of air is completely meaningless, as any temperature difference drives convection. That's why it's very wrong to say that air is "thermal insulator", and assume that any potting compound would be automagically better. For example, laptop power supplies do great job of dissipating a lot of power within small frame, and they do not do that by potting, they use heat spreaders (metal or pyrolytic graphite, which is 3-5 times better than pure copper, but only works as thin sheets) to couple that heat over larger area of internal air and external plastic case, both pretty poor thermal conductors, but given enough cross-sectional area, it works out, therefore the initial heat spreading is the key to success.

Ah this guy again. Even though I just finished several dozens of thermal measurements with potting compounds, he knows better. I had enough of you.

[coppercone2]

it entirely depends on design of the PCB, enclosure and ratio of potting compound to enclosure............ I don't see how dozens measurements can change this if you don't have the specific board layout and enclosure that this effect shows up on.


Don't be a pot salesman

why do I doubt that you setup dozens of experiments with pyrolytic graphite sheet and advanced thermal design to correctly use that sheet?