THERMAL DESIGN

Let's begin by defining the main terms. The goal of thermal management is to control the temperatures of the electronic devices. But what is

temperature

? Technically speaking, on a microscopic level it is a measure of the average molecular kinetic energy in the matter. The normal flow of kinetic energy is from a higher temperature region (or an object) toward a lower temperature region (or an object). This flow is referred to as heat transfer.

In general, there are three types of heat transfer: conduction, convection, and radiation. Conduction is the collisional transfer of energy between atoms, which occurs in solids. Convection is the motion of molecules in air or fluids. Radiation is the energy flow by electromagnetic waves. In practical electronics primarily only the first two types are usually taken into account.

Why do electronic devices get hot? Electric currents and alternating electromagnetic fields cause power dissipation in all electronic parts, which results in increase of their temperatures. This in turn affects the reliability and life expectancy of these components. The measures of the electronic system reliability are failure rate and its inverse, mean time between failures (MTBF). According to Arrhenius model, each 10^oC rise increases the failure rate by 50%.

At certain point any electronic device can be irreversibly destroyed. Typical maximum operating temperature "T" for semiconductors and ICs is 125-175 ^oC at their junctions, capacitors- 85-125 ^oC, wire insulation- 105-200 ^oC. The thermal management and engineering whose task is to control "T" of the product, is therefore an essential part of electronics design.

When designing a heat sink for a semiconductor cooling, first of all you need to choose the component's maximum operating junction temperature Tjmax (typically, we select 105-120 ^oC for commercial parts). Then, for convection cooling the required heatsink thermal resistance should be
R_th-hs<(Tjmax-Ta)/P-(R_thj-c+R_thc-hs) ^oC/W, where Ta- ambient, R_thj-c - thermal resistance between junction and the case from the datasheet (typically 0.5-2.5 ^oC/W for conventional discrete power packages), R_thc-hs - thermal resistance between the device's case and the heatsink, P- power dissipated by the device in watts. Besides conduction cooling, there is another path for the heat- from the case of the device directly to the air. This path is defined by junction to ambient thermal resistance R_thj-amb. However, you can usually disregard it when a device is mount onto a heatsink.

See thermal circuit diagram to the left for an illustration. Once you found the required R_th-hs, you can pick a stamped or extruded heat sink with equal or lower value of thermal resistance. Note that if you use an insulator between the device and the heatsink, you need to take it into account as well. For off-the-shelf parts, Rth-hs is normally specified in the datasheet. Harry Lythall found empirically a "rule of thumb" calculation formula for R_th-hs of home made U-shape folded aluminum sheet. Based on his equation, the required surface area in sq.cm is A=(50/Rth-hs)².
In general, the main optimization criteria are to maximize the exposed heat exchanger's surface area, and to minimize its weight and the mean distance of the exposed surface from the component to be cooled.

Below you will find free calculators and useful information on the thermal design.

Disclaimer, Disclosure and Terms of Use | Contact Us | About | Privacy

© 2004, 2008-2015 Lazar Rozenblat