1
\$\begingroup\$

I found myself needing to attach small heatsinks to Mosfets and CC power regulators. Also some high-power diodes.

Today I have learnt that thermal compounds, even best in the market, are dozens of times less conductive than aluminium or copper.

I understand thermal compound is used only to cover microscopic cracks.

But what about thermal adhesive tapes?

My question is:

Using a thermal tape, I might completely cover the surface between the heat source and the heat sink in order to join them together.

If thermal tape is 100 times less conductive than aluminium, how do they manage to do their job and transfer the heat even at high power applications?

\$\endgroup\$
4
  • 5
    \$\begingroup\$ they are very thin. \$\endgroup\$
    – analogsystemsrf
    Aug 24, 2019 at 3:42
  • \$\begingroup\$ Thin is always worse when it comes to thermal. \$\endgroup\$
    – Voltage Spike ♦
    Aug 24, 2019 at 5:48
  • 3
    \$\begingroup\$ @VoltageSpike if that was true, we could insulate houses by making the walls very thin. \$\endgroup\$
    – rackandboneman
    Aug 24, 2019 at 23:17
  • 2
    \$\begingroup\$ @rackandboneman @#^! I always get conductivity confused with resistivity, its a terrible problem \$\endgroup\$
    – Voltage Spike ♦
    Aug 25, 2019 at 1:54

4 Answers 4

Reset to default
6
\$\begingroup\$

The total thermal resistance is the length * resistivity / Ac, where length is the length of the conductive path, Ac is the cross-sectional area of the conductive path, and resistivity is the thermal resistivity of the pad.

The conductive path length is actually just the thickness of the pad. So even though the resistivity may be high, because the thickness of the pad is very small, it is an OK thermal conductor overall.

It is a design goal to make the pads out of the best all around material. It is hard to find anything conformable that can act as a heat sink pad that is extremely conductive like copper or aluminum. Sometimes the pads also need to provide electrical insulation.

\$\endgroup\$
2
  • 2
    \$\begingroup\$ Thank you for having read my question. :) So because the tape is thin, heat does manage to pass through, and despite being 100x less conductive than metal, it is a non-issue as barrier is small. Thanks! \$\endgroup\$
    – Phil
    Aug 24, 2019 at 10:49
  • \$\begingroup\$ @marcelm definitely. Thanks. \$\endgroup\$
    – user57037
    Aug 25, 2019 at 0:15
6
\$\begingroup\$

Heat sinking compounds/tapes are less conductive than metals but are orders of magnitude better than air. The main goal of these interfaces is to fill in the gaps between the solid interface between your heat source and sink. Without such compounds air pockets would form and the thermal interface would be poor. See the image below (source). The compound takes up any space between rough interfaces. Compare thermal interface options using the table at end of description and you will see that the compounds make sense.

enter image description here

A single solid piece of copper or aluminum mountain directly to the chip die would certainly heat sink better but the issue is it would short out all the circuitry, thus epoxy encapsulation must be used for an isolating interface and as a result then some compound to interface to a metal heat sink.

When cost is a non-issue even more novel thermal interfaces can be developed. I have head of Chemical Vapor Deposition diamond used in certain packages as the diamond is superior to an epoxy interface but still has insulating properties. Example source, tho I have never personally used such devices.

Thermal Conductivity Table:

Air Thermal Conductivity: ~26mW/mK

EMC Packacging Compound: ~ 2-4W/mK

Thermal Compound Thermal Conductivity High Grade: 8.5 W/mK

Aluminum: 205 W/mK

Copper: 385 W/mK

Diamond: 2200 W/mK

\$\endgroup\$
3
  • \$\begingroup\$ Hello. Thank you for your answer. However as I tried to explain, I do understand the purpose and magic of thermal compounds. My confusion comes from the fact that, whereas, thermal liquid compound can be applied in thin amounts and made to fill the gaps by pressure, a thermal tape covers (or may cover) 100% of the surface, this creating a blocking layer between 2 metals. If these 2 metals have 100x or 200x more thermal conductivity, how is it that thermal tape, blocking the contact between the two, with its 100x less conductivity, still does the job please? \$\endgroup\$
    – Phil
    Aug 24, 2019 at 10:47
  • \$\begingroup\$ @Phil. Let's do a thought experiment. Imagine 2 copper surfaces that after machining will only have direct contact on 1% of surface area when pressed together. The remaining 99% are voids that don't touch directly. Now you can leave the remaining 99% area interface to work through air (bad conductivity) or something else like thermal compound (good conductivity). Your effective thermal conductivity is 1% copper interface + 99% whatever else you chose. You are better off choosing thermal compound as total effective would be lower than air. \$\endgroup\$
    – EasyOhm
    Aug 24, 2019 at 18:29
  • \$\begingroup\$ @Phil. Similar thing for tapes. A thin layer of tape will interrupt the 1% of superior copper to copper contact but will certainly improve the remaining 99% that would have been air. The end total conductivity will go down. You have to imagine that things are actually usually very rough at small scales and very few things actually touch. \$\endgroup\$
    – EasyOhm
    Aug 24, 2019 at 18:31
5
\$\begingroup\$

If thermal tape is 100 times less conductive than aluminium, how do they manage to do their job and transfer the heat even at high power applications?

They do their jobs because they are better than air, which is lower than 0.024 W/(m*K) and so they are approximately 100 times better than air.

If your using an adhesive or thermal compound, it needs to be thin. The thickness also affects the thermal conductivity. The total conductivity will go down when the thickness goes down. (Think of Teflon pans, Teflon is one of the least thermally conductive materials, but if applied in a thin layer it still conducts enough heat to heat food) Many adhesives have thicknesses in the teens or hundreds of um.

You also need a 'filler' in between two pieces of metal, as they are not perfectly flat (the more flat you want metal the higher the machining cost). This allows air to come between two metal surfaces, again air has a low conductivity, and adhesive is much higher than air.

By the way, there are new adhesive materials made of graphite or graphene, that have conductivities in the 400 W/(mK) to 1000 W/(mK) (in the x-y direction) that you might want to check out.

https://industrial.panasonic.com/ww/products/thermal-solutions/graphite-sheet-pgs/pgs

You can get pads at major distributors

\$\endgroup\$
3
  • \$\begingroup\$ Yup graphite is really good! X-y great at spreading the heat. It is a pig to deal with though as it tears soo easily \$\endgroup\$
    – user16222
    Aug 24, 2019 at 7:14
  • \$\begingroup\$ Yes but air or thermal compound get only into the cracks. Where the surface is smooth, they touch without the compound. My confusion comes from the fact that, whereas, thermal liquid compound can be applied in thin amounts and made to fill the gaps by pressure, a thermal tape covers (or may cover) 100% of the surface, this creating a blocking layer between 2 metals. If these 2 metals have 100x or 200x more thermal conductivity, how is it that thermal tape, blocking the contact between the two, with its 100x less conductivity, still does the job please? \$\endgroup\$
    – Phil
    Aug 24, 2019 at 10:47
  • \$\begingroup\$ Edited the question \$\endgroup\$
    – Voltage Spike ♦
    Aug 24, 2019 at 14:55
2
\$\begingroup\$

How do they manage to do their job...

I am more familiar with pads than adhesive tape, but they are similar. Although they are not as thermally conductive as metal, they are more thermally conductive than air. They assure good physical contact, and they are very thin.

A typical thermal pad has a thermal resistance of 1 W/mK and a thickness of 0.2 mm. One square cm of this pad has a thermal resistance of 0.5 degC/W. Not bad.

\$\endgroup\$
2
  • \$\begingroup\$ Thank you for your kind answer. What I don't understand is, tapes cover the whole area whereas compound can fill small cracks and small cracks only. So if tape/pad covers the whole area with 100x less conductivity, how does it work? \$\endgroup\$
    – Phil
    Aug 24, 2019 at 10:48
  • \$\begingroup\$ The tapes are also referred to as "gap -filler". They are very soft and will confirm to the surface texture of the two surfaces to fill the air gaps. They are used because thermal compound is sticky and difficult to remove from surfaces. In a production environment where you're building multiple units, tape which stays in place is much easier to deal with. In a prototype environment where you're testing systems, tape is easier to deal with, as you don't want thermal compound all over your hands, bench, tools, and clothes. \$\endgroup\$
    – rfdave
    Aug 24, 2019 at 12:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

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