1
\$\begingroup\$

A pcb in deionized water may behave differently due to the change in capacitive ratio (leaving obvious problems such as where to get pure deionized water and keep it contaminant free).

However, would covering the circuit with thermal paste have the same effect?

From my understanding, thermal paste is thermally conducting but a good electrical insulator. From basic physics, heat dissipation works better with larger surface area (and a large temperature gradient between two sources, amongst other factors such as materials used etc.

Why is this not used in complex high power circuits where passive power dissipation is required?

Would it be a problem if it was applied only across the copper tracks and on the tops of components, making an irregular heatsink that spans the dimensions of the board?

Obvious use case would be cramped Mobile devices: This would also increase the surface area, which would increase heat dissipation efficiency significantly.

\$\endgroup\$
  • 3
    \$\begingroup\$ This should really be on super-user not EE.SE. But either way thermal paste is not a good thermal conductor in general. It's fine in a very thin layer (for filling in roughness of a thermal surface), but isn't some magic cooling paste. \$\endgroup\$ – Tom Carpenter Jun 15 '16 at 0:09
  • \$\begingroup\$ Finally, a large surface area only helps if it is as a contact area with whatever it is dissipating the heat in to (i.e. the air). The fact that it is lots of small particles makes it worse for heat transfer compared with a solid block of metal. \$\endgroup\$ – Tom Carpenter Jun 15 '16 at 0:17
  • \$\begingroup\$ @ThePhoton technically it is but, I concede that the comment was not a complete picture. Gone now. \$\endgroup\$ – Tom Carpenter Jun 15 '16 at 2:03
  • \$\begingroup\$ One of the most effective transport mechanism for thermal energy is convection (moving heated matter). Often it is more effective to transport heat using convection in a medium of low thermal capacity (air) than using thermal conduction in a medium of quite good thermal conductivity. So I assume covering the whole circuit with thermal paste that can not do convection (because it is so gooey) might even be much worse than leaving the circuit in direct contact with air that can convect. \$\endgroup\$ – Curd Jun 15 '16 at 6:45
  • 1
    \$\begingroup\$ BTW: not having air around the circuit that can transport the heat away from the components is a special challenge for electronics used in space. There you can rely only on heat transfer by conduction. \$\endgroup\$ – Curd Jun 15 '16 at 6:52
3
\$\begingroup\$

PC boards are designed with traces wide (and thick) enough so that they DO NOT dissipate any significant amount of power. Therefore they do not require heat sinking, by water (or any other liquid) immersion, nor by mechanical heat sinking. A PC board that heats up under normal use would be poorly designed by definition.

Thermal paste DOES NOT dissipate heat. It is only a medium of contact exchange between the heat-generating object (typically a semiconductor or resistor) and the heat sinking device (an air-cooled heat sink or liquid-cooled heat exchanger).

For those reasons you don't see PC boards covered in thermal paste.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.