A GED is a bad thing

New service: GED heating expertise

Adequate thermal management is part of practically every electronics development. If this point is underestimated or recognized too late, there is a risk of extensive reworking, delays and additional costs. GED has therefore set up a comprehensive and efficient service for all aspects of the calculation and simulation of the heat generation in electronic circuits.

For every electrical assembly and every electrical equipment there is a certain temperature range in which its functionality is guaranteed. The lower limit value can always - provided there is an adequate supply of energy - be guaranteed simply by heating. Alternatively, it can be maintained with sufficient insulation from a cold environment due to the component's own heat loss or the operating equipment. However, it is precisely this heat loss that causes problems if the upper limit value is reached. Compared to heating that converts various forms of energy into heat, there is no effect that could convert heat into another form of energy and thus render it harmless. Only the independent heat transport from warm to cold and the heat transport from cold to warm with the aid of additional energy are possible. The latter solution is rarely used in electronics cooling. In almost all cases of the analysis of electronics cooling, the heat transfer from warm to cold is considered. This consists of a heat source, a heat path and a heat sink, consisting of a fixed ambient temperature.

It is therefore of great importance first of all how strong the heat source is in the assembly to be examined. Since mostly electrical loss effects occur, a network analysis with calculation of the power losses can provide this value or these values.

Linear networks can usually be determined analytically with regard to their power loss with reasonable effort. Nonlinear ones - which include almost all modern power electronic circuits - require simulation in a numerical software tool or simulation software. Graphic SPICE simulators are available on the market and used by GED, which contain the most common component models or in which models of specific components can usually be easily entered.

Since electrical components usually have strongly temperature-dependent properties, which in turn influence their power loss and ultimately also their heating, it is possible and necessary to take into account the temperature behavior of the components in the numerical simulation. Each component model has coupled electrical and thermal properties as well as connections for the electrical network such as a thermal cooling network. This is a rough approach, since the cooling network can only be designed one-dimensional; however, it is a good instrument for examining the stability of a circuit. For example, the RdsOn of a MOSFET is highly temperature-dependent; in an electricity-driven regime this can be fatal; an RdsOn that rises with the temperature results in a higher power loss, thus a rising temperature and, in turn, a higher RdsOn.

Calculation and simulation of one-dimensional heat paths

Heat transport and electricity have many analogies; With the exception of inductance, there is a thermal counterpart for every basic component in electrical engineering: the power source becomes a heat source, the voltage source becomes a fixed ambient temperature or heat sink, the resistance becomes a thermal resistance and the capacity becomes a heat capacity.

With these basic components, simple heat paths - for example the arrangement Die ?? - component housing-heat sink-environment - can be analyzed. When using a numerical simulation program, complex load profiles can also be viewed.

The effect of thermal radiation is non-linear; this can ultimately only be calculated as a non-linear thermal resistance with numerically operating software tools.

Simulation of thermal paths at the assembly level

For the calculation methods mentioned so far, sometimes complex modeling is a prerequisite. However, GED has a software tool that automatically generates a thermal model of an assembly from production data. Compared to complex 3D flow simulation tools, the calculation of the physical effects is modeled simply and efficiently. It is a conservative calculation method; Calculated temperature increases turn out to be greater in the simulation than in reality. This means a calculation on the safe side. This is sufficient for an initial assessment of the heating of an assembly during operation.

The software considers the heating of the circuit board due to the conduction losses in the copper as well as the heat input from components. However, load profiles cannot be simulated; the currents are considered to be direct currents. The rms values ​​of the currents must therefore be known for a calculation. Since the program requires a current sum of 0 in every energized network, you have to weigh up which currents are set in the simulation, since the rms values ​​of alternating currents in a network usually do not have the sum 0.

The basis for the thermal simulation is a current density simulation in the copper line. This alone already provides information on how the design can be adapted in order to optimally lay out surfaces and conductor tracks. The simple data transfer from all common layout programs offers the possibility to run through optimization loops quickly.



The steady thermal state of the assembly is calculated, i.e. after all equalization processes have subsided. Here, the heat output per surface element on the assembly is calculated against a heat dissipation constant with the dimension W / m²K. A calculation tool in the software enables this value to be determined for the most common installation situations of assemblies.

Even if absolute temperatures cannot be calculated, it is possible to identify hotspots on the assembly and thus adapt the design accordingly - be it by changing the conductor pattern, the layer structure or the component arrangement.

In Figure 2, a bottleneck in the current density distribution can be seen in point 1, which, however, as shown in Figure 3, does not lead to a strong increase in temperature. In Figure 3 at point 2, a strong increase in temperature can be observed. It is mainly due to the comparatively poor thermal connection of the power transistor at this point. Lower peak temperatures and better heat spread can be seen at point 3 in Figure 3; here the power transistor is fully connected and there is a lot of copper in the inner layers.

These are simple examples on a specially created test layout in which effects are provoked by the design. But GED also has modern, sophisticated tools for most of the other layouts to identify bottlenecks and hotspots and to identify possible remedies. This enables GED to provide its customers with comprehensive support early on in terms of thermal management, even during development. Possible errors that can later only be eliminated with great effort are avoided right from the start - quickly and with little effort.

Would you like to find out more about heat expertise and heat management à la GED? Talk to us:

Call us:

+ 49 (0) 2247 92 19-0.

Or send one message.