Power Converter Thermal Impedance – A Guide to the Essentials
May 27, 2020
All power converters dissipate power internally as heat and so run warmer than their surroundings. As long as this additional warmth can be transferred away to the surroundings, without critical internal temperature limits being exceeded, the converter can run at full power
At a certain ambient temperature, the converter reaches its maximum internal temperature limit and any further increases in ambient temperature must be compensated for by reducing the amount of power dissipated inside the converter by reducing the load. This is called thermal derating1 .
The thermal impedance of a converter from case to ambient can be found from the following relationship:
So, if the power dissipation is known (the difference between the input power and the output power), then measuring the case temperature rise above ambient allows the thermal impedance to be determined.
In practice, this simple relationship is very hard to determine. Measuring the converter’s temperature rise above ambient has to be done with a very high precision to get reliable results. One way of increasing the reliability of the result is make several measurements at different ambient temperatures as the ΔT (temperature rise above ambient temperature) should be the same for each temperature step.
However, for true convection cooling operation, the thermal impedance has to also be determined in a draught-free environment, as the air movement caused by the thermal chamber’s fan can influence the result. If the device-under-test (DUT) is placed inside a cardboard box inside the temperature-controlled chamber, and a four-wire (Kelvin contact) system for terminal voltages is used in power measurements, then reasonably accurate measurements can be made:
The wire mesh grid allows air movement while blocking draughts
Fig. 1: Thermal impedance test set up (DC/DC) using a thermal chamber
The reason why the temperature measurements need to be carried out in a draught-free environment is because the thermal impedance changes drastically under different conditions – it is very dependent on the efficiency with which heat is transferred to the surrounding fluid (in most cases, this ‘fluid’ is air). If the heat transfer coefficient is poor, then the dissipated power will cause a greater increase in the case temperature and the thermal impedance will be higher. If the heat transfer is more efficient, then the temperature rise will be lower for the same dissipated power – a low thermal impedance. So, equation 1 is correct, but only under one particular set of conditions (convection cooling only). It is also important that the testing conditions stay constant to give reproducible results – the documentation of the set-up, the list of equipment used and the calibrations carried out are just as important as the test result itself.
For most real-life applications, the convection cooling thermal impedance figure is the most useful because this corresponds to the way that PCB-mounted converters are typically used. They are soldered onto a circuit board and the heat dissipated inside the converter is then spread out into the surroundings by free air convection cooling. Some heat will be also conducted away via the electrical connections to the PCB tracks, so some temperature derating graphs given in the datasheets are accurate only if the PCB layout information is also followed. A small amount of heat will be additionally dissipated by radiation to the surroundings, but in our experience, this effect is minor.
The most effective way of increasing the efficiency of heat transfer to the surroundings is to move the fluid about (this is why the fan in the thermal chamber can affect the readings). This effect has been known about since Isaac Newton first published a paper in 1701 containing the relationship:
Where q is the rate of heat transfer, h is the coefficient of heat transfer, A is the surface area of the DUT and ΔT is the temperature difference between the DUT and the ambient.
Newton’s Law of Cooling states that for a given temperature difference, the rate of heat transfer can be increased by increasing the surface area (for example, by adding a finned heatsink to the DUT) or by improving the heat transfer coefficient (for example, by blowing air over the hot part).
In practice, the heat transfer coefficient typically changes abruptly at a boundary condition when the air flow over the converter transitions from laminar to turbulent flow. For most power converters with flat plate upper surfaces, this transition point is around 0.1-0.2m/s. Thus, an airflow of 0.1m/s (20LFM) can be considered to be convection cooled only and anything higher can be considered to be forced air cooling.
Fig. 2: Thermal impedance vs. wind speed
RECOM has its own in-house wind tunnel to accurately measure the thermal impedance of our products with forced cooling. The air flow inside the wind tunnel is laminar thanks to a honeycomb flow conditioner element and the output diffuser element eliminates back pressure variations, ensuring an even pressure and airflow profile in the central test volume.
A precision air flow sensor connected to a feedback circuit driving the fan guarantees a stable and accurately controllable air flow. The device temperature is measured using a thermal camera to avoid any turbulence effects caused by inserting a foreign object into the air flow close to the DUT.
Fig. 3: RECOM’s in-house wind tunnel
The small circular window is made from a special infra-red transparent glass to permit remote IR camera temperature monitoring of the device-under-test. The precision flow sensor top left is connected to the fan control unit on the right to accurately regulate the air flow.
With such equipment, we are able to accurately test and measure the convection cooling and forced cooling parameters of our products for our datasheets, for example, the RPA200H:
We are here to help
RECOM prides itself on designing compact, cost-effective and above all highly efficient AC/DC, DC/DC and switching regulator power supplies. The efficiency of a converter, in particular a PCB-mount module where the power density can be very high, is often a critical deciding factor when choosing the most suitable part for a particular application. For example, a converter with 96% efficiency will have half the internal dissipated heat compared to a converter with 92% efficiency.
We carry out exhaustive testing using our in-house automated test system, our climate chambers and using our wind tunnel to provide accurate temperature-based data and information for our datasheets, such as thermal impedance and maximum case temperature figures, efficiency vs. load tables and temperature derating graphs for both standard parts and those that have a heatsink option.
We do this to allow our customers to select and design-in the optimum part for their application so that they can be confident that the converter will function within its performance limits at both extremely cold (down to -40°C) and very hot (up to 100°C) ambient temperature environments. As the choice of converter and the cooling system used can be critical to the success of a project, it is worthwhile contacting RECOM technical support or one of our experienced sales engineers for advice.
RECOM: We Power your Products
1, Refer to the Recom DC/DC Book of Knowledge), Chapter 3 for more information.
The thermal impedance of a converter from case to ambient can be found from the following relationship:
| Eq. 1: |  |
|---|
So, if the power dissipation is known (the difference between the input power and the output power), then measuring the case temperature rise above ambient allows the thermal impedance to be determined.
In practice, this simple relationship is very hard to determine. Measuring the converter’s temperature rise above ambient has to be done with a very high precision to get reliable results. One way of increasing the reliability of the result is make several measurements at different ambient temperatures as the ΔT (temperature rise above ambient temperature) should be the same for each temperature step.
However, for true convection cooling operation, the thermal impedance has to also be determined in a draught-free environment, as the air movement caused by the thermal chamber’s fan can influence the result. If the device-under-test (DUT) is placed inside a cardboard box inside the temperature-controlled chamber, and a four-wire (Kelvin contact) system for terminal voltages is used in power measurements, then reasonably accurate measurements can be made:
The wire mesh grid allows air movement while blocking draughts
Fig. 1: Thermal impedance test set up (DC/DC) using a thermal chamber
The reason why the temperature measurements need to be carried out in a draught-free environment is because the thermal impedance changes drastically under different conditions – it is very dependent on the efficiency with which heat is transferred to the surrounding fluid (in most cases, this ‘fluid’ is air). If the heat transfer coefficient is poor, then the dissipated power will cause a greater increase in the case temperature and the thermal impedance will be higher. If the heat transfer is more efficient, then the temperature rise will be lower for the same dissipated power – a low thermal impedance. So, equation 1 is correct, but only under one particular set of conditions (convection cooling only). It is also important that the testing conditions stay constant to give reproducible results – the documentation of the set-up, the list of equipment used and the calibrations carried out are just as important as the test result itself.
For most real-life applications, the convection cooling thermal impedance figure is the most useful because this corresponds to the way that PCB-mounted converters are typically used. They are soldered onto a circuit board and the heat dissipated inside the converter is then spread out into the surroundings by free air convection cooling. Some heat will be also conducted away via the electrical connections to the PCB tracks, so some temperature derating graphs given in the datasheets are accurate only if the PCB layout information is also followed. A small amount of heat will be additionally dissipated by radiation to the surroundings, but in our experience, this effect is minor.
The most effective way of increasing the efficiency of heat transfer to the surroundings is to move the fluid about (this is why the fan in the thermal chamber can affect the readings). This effect has been known about since Isaac Newton first published a paper in 1701 containing the relationship:
| Eq. 2: |  |
|---|
Where q is the rate of heat transfer, h is the coefficient of heat transfer, A is the surface area of the DUT and ΔT is the temperature difference between the DUT and the ambient.
Newton’s Law of Cooling states that for a given temperature difference, the rate of heat transfer can be increased by increasing the surface area (for example, by adding a finned heatsink to the DUT) or by improving the heat transfer coefficient (for example, by blowing air over the hot part).
In practice, the heat transfer coefficient typically changes abruptly at a boundary condition when the air flow over the converter transitions from laminar to turbulent flow. For most power converters with flat plate upper surfaces, this transition point is around 0.1-0.2m/s. Thus, an airflow of 0.1m/s (20LFM) can be considered to be convection cooled only and anything higher can be considered to be forced air cooling.
Fig. 2: Thermal impedance vs. wind speed
RECOM has its own in-house wind tunnel to accurately measure the thermal impedance of our products with forced cooling. The air flow inside the wind tunnel is laminar thanks to a honeycomb flow conditioner element and the output diffuser element eliminates back pressure variations, ensuring an even pressure and airflow profile in the central test volume.
A precision air flow sensor connected to a feedback circuit driving the fan guarantees a stable and accurately controllable air flow. The device temperature is measured using a thermal camera to avoid any turbulence effects caused by inserting a foreign object into the air flow close to the DUT.
Fig. 3: RECOM’s in-house wind tunnel
The small circular window is made from a special infra-red transparent glass to permit remote IR camera temperature monitoring of the device-under-test. The precision flow sensor top left is connected to the fan control unit on the right to accurately regulate the air flow.
With such equipment, we are able to accurately test and measure the convection cooling and forced cooling parameters of our products for our datasheets, for example, the RPA200H:
We are here to help
RECOM prides itself on designing compact, cost-effective and above all highly efficient AC/DC, DC/DC and switching regulator power supplies. The efficiency of a converter, in particular a PCB-mount module where the power density can be very high, is often a critical deciding factor when choosing the most suitable part for a particular application. For example, a converter with 96% efficiency will have half the internal dissipated heat compared to a converter with 92% efficiency.
We carry out exhaustive testing using our in-house automated test system, our climate chambers and using our wind tunnel to provide accurate temperature-based data and information for our datasheets, such as thermal impedance and maximum case temperature figures, efficiency vs. load tables and temperature derating graphs for both standard parts and those that have a heatsink option.
We do this to allow our customers to select and design-in the optimum part for their application so that they can be confident that the converter will function within its performance limits at both extremely cold (down to -40°C) and very hot (up to 100°C) ambient temperature environments. As the choice of converter and the cooling system used can be critical to the success of a project, it is worthwhile contacting RECOM technical support or one of our experienced sales engineers for advice.
RECOM: We Power your Products
1, Refer to the Recom DC/DC Book of Knowledge), Chapter 3 for more information.
Applications
