Technical Information Site of Power Supply Design

• Basic Knowledge
• Thermal Design
• Thermal Design of Semiconductor Components in Electronic Equipment
• Thermal Resistance Data: θJA and ΨJT in Estimation of TJ: Part 2

2021.12.22 Thermal Design

# Thermal Resistance Data: θJA and ΨJT in Estimation of TJ: Part 2

Thermal Design of Semiconductor Components in Electronic Equipment

In the previous article, as Part 1 of a discussion of how θJA and ΨJT are or can be used in calculations to estimate TJ, it was explained what can be done using θJA and ΨJT. In this Part 2, the characteristics of ΨJT and the usefulness of θJA and ΨJT in estimating TJ are explained. As also indicated in the previous article, a calculation example for TJ estimation using thermal resistance data will be presented separately.

## Condition-Dependent Characteristics and Usefulness of θJA and ΨJT

In the previous article it was explained that ΨJT is a thermal characterization parameter representing the temperature difference between the junction and the center of the top surface of the package relative to the overall device power consumption, and so can be utilized in estimates of TJ in an actual operating state in actual equipment. However, in an actual operating state, there are conditions that affect θJA and ΨJT, and that alter the usefulness in estimation of TJ. Here a number of examples are considered, and their respective characteristics and the consequences for usefulness are considered.

●Changes in the Heat Dissipation Performance of a PCB
The graph on the right indicates the relation of the PCB surface copper foil area to θJA and ΨJT. Whereas θJA is greatly influenced by the copper foil area, that is, the heat flowing into the PCB, where ΨJT is concerned, because most of the device heat flows into the PCB, the TJ-TT temperature difference is extremely small, so that the value of and changes in ΨJT are also small. Hence θJA changes considerably depending on the conditions of the PCB, making it difficult to use θJA without modification in estimations of TJ; but ΨJT does not change much for different PCBs, and so can be utilized.

●State of Being Covered by a Shield Case etc.
Due to EMC considerations and the like, a target device may be covered by a shield case. Below, actual measurements of θJA and ΨJT with and without shield cases are compared.

When a shield case is present, both θJA and ΨJT are higher, but fluctuations in θJA are large, and it cannot be used in estimation of TJ. The increase in ΨJT is slight, and the value is itself small with only small changes, so that use as-is will not result in large errors when estimating TJ. As an example, when using ΨJT = 9.4°C/W without a shield case, calculating TJ for a device with a shield case, compared with an actual temperature of 106.7°C, the error is less than 1%.

●State of Being Resin-Sealed and Covered by a Shield Case etc.
In some cases, a PCB may be resin-sealed with mounted components for protection purposes. Here, a state in which a target device is further covered by a shield case or the like is considered.

Resin sealing causes thermal resistance to decrease dramatically, and under these conditions, θJA cannot be used in estimation of TJ. ΨJT tends to increase and fluctuates considerably. Using the value ΨJT = 9.4°C/W when there is no resin sealing or shield case, TJ for a device with resin sealing and a shield case is calculated as follows.

Compared with the actual temperature of 53.3°C, the error is about 8%. Study is necessary to determine whether to allow this error or to correct for it when calculating TJ.

●State in Which Heat-generating Devices are Adjacent to Each Other
While this is something that should be avoided, there are cases in which heat-generating devices are adjacent. Below are θJA and ΨJT values when two devices are at an appropriate distance (center) and when they are adjacent (right).

As the data indicates, when the devices are close to each other, both θJA and ΨJT increase, but in the case of θJA the change is large, so that it cannot be used in estimating TJ. Of course ΨJT also changes, but the value was initially small and the change is also small, so that using it to estimate TJ does not result in a large error. As an example, using the value ΨJT = 9.4°C/W for a single device, the TJ in a state in which two devices are adjacent is calculated as follows; compared with the actual temperature of 101.5°C, the error is less than 1%.

●Case in Which the Number of Layers in the PCB Changes
The way θJA and ΨJT change when the number of layers in the PCB changes is indicated below.

When the number of layers in the PCB increases, θJA drops dramatically; in this case also, θJA cannot be used to calculate TJ. ΨJT likewise changes considerably, but if, as an example, the value ΨJT = 9.4°C/W for 1s (a single-layer PCB) is used to calculate TJ for the case of 2s2p (a four-layer PCB), we obtain the following.

Compared with the actual temperature of 36.3°C, the error is about 6%. Whether this can be allowed or correction must be performed is a matter for study.

## Summary

We have compared θJA and ΨJT on the assumption of installation in actual equipment under four different conditions. As in the previous article, θJA is affected by such heat dissipation conditions as the copper foil and the PCB as well as thermal interference from adjacent components, and so we find that it cannot easily be used in calculations to estimate TJ in actual equipment.

On the other hand, while ΨJT also changes with the mounting conditions, its value is originally small, and the value changes only slightly for different conditions. By ascertaining the PCB to be used and the state of the actual hardware and by precisely measuring TT, it is possible to use a ΨJT value that has been provided in estimation of TJ in actual use.

#### Key Points:

・It is essentially not possible to estimate TJ using θJA while the device is installed in actual equipment.

・ΨJT is different depending on the mounting conditions, but with an understanding of the PCB being used and the mounted state, it can be used for TJ estimation with the device installed in actual equipment.