One of the steps, when determining the various elements constituting a power converter, is the choice of the dissipator, in order to preserve the thermal integrity of the power semiconductors. This is achieved by keeping the junction temperature of the component below its critical value during the operating cycle.
The cost of the heatsink, or more generally of the cooling function, is closely linked to the heatsink-component couple.
For example, to keep one or two TO3 enclosures at temperature, the cost of the “dissipator + fan” couple is often greater than the sum of the costs of the components to be cooled. As a general rule, the higher the power of the application, the more the “semiconductor + control stages” couple is preponderant, facing the dissipator station. This is partly due to the cost of power semiconductors.
The choice of heatsink can only be made with knowledge of the following elements:
-
number and type of components to be cooled, therefore knowledge of the case used for each component (discrete component, module, presspack, etc.);
-
losses generated by each component (depending on the operating cycle and assembly topology), possible overloads;
-
desired cooling mode (natural convection, forced ventilation, water chambers, etc.);
-
mechanical stresses and integration of the system in its final environment (dielectric stresses, assembly fixing, enclosure).
The purpose of this dossier is to develop the industrial approach for choosing a heatsink, a compromise between academic calculation and sometimes laborious simulation.