Thermal circuits model heat flow from the power supply to the outside world (see figure 1). Every power supply, no matter how efficient will generate some heat from the power it dissipates (Pdiss). Heat flow(Q) behaves in a similar manner as electrical current. In the thermal circuit, a voltage source is remodeled as a heat source. Each heat source within a power supply will encounter thermal resistances Rth (θ) which in turn cause thermal drops that restrict the flow of heat. The thermal profile of a power supply is therefore measured by the aggregate Rth (θ) expressed as Cº/W, or how many degrees (Celsius) are generated per 1watt of dissipated power. A properly modeled thermal circuit illustrates the effectiveness of a thermal design or where improvement is needed.
Figure 1: Thermal Circuit
The focal point of any thermal design is to determine the maximum ambient operating temperature (Ta max). The supply's Ta max must not cause a thermal failure of any individual component under nominal continuous rated load conditions. If the Ta max does not meet the desired system requirement, the thermal designer can employ several cooling methods to manage the thermal profile.
They're explained briefly here.
A. Convection: Still air or through forced air flow by fans/blowers. Air is the traditional surrounding environment for convection cooling, but liquid cooling is also used.
Cincon's CFM300M http://psui.com/series/cincon/cfm300m
Heat transferred from one solid material to another generally larger mass i.e heat sink. Heat sinks and other transfer mediums have a thermal conductivity rating.
Absopulse's PPF300 http://psui.com/series/absopulse/ppf-300
Power supplies mounted in enclosures will radiate heat to the enclosure, but also receive heat from local dissipative elements. This makes it hard to quantify or model the radiant properties of power supplies. Not often used.
Several cooling methods may be applied simultaneously.
A good thermal design should apply techniques that show a measured reduction in thermal resistance Rth (θ). This allows for a higher Ta max given the same power dissipation.
In designs where applied cooling methods cannot provide a safe thermal environment, output power must be reduced. Derating allows for higher operating temperatures by reducing output power thereby reducing power dissipation at those temperatures. Power supply manufacturers specify a derating curve for the amount of load that can be applied versus temperature (see figure 2).
Figure 2: Typical Derating Curves
Ambient operating temperatures Ta should be measured, since the actual temperature could be higher than anticipated due to localized heating, including that of the supply itself.
• Where only convection cooling is possible, allow for sufficient space around the converter for heat escape and circulation.
• Operate with derated output power if possible.
• Operate with additional cooling if possible.
• Consider the effects of temperature on power supply reliability and lifetime.MTBF, or Mean Time Between Failures is an accepted measure of reliability. Temperature is a key factor in the MTBF calculation.
In summary, good thermal practice requires a power supply to operate within a specified temperature with a high degree of reliability.
By John Benatti