# How to best relate a case temperature measurement with the actual junction temperature?

If I have 4W of power dissipated through a transistor (as derived from measured current and the datasheet's on-resistance) and I measure 40 deg. C on the case of the transistor, what is the best way to calculate its junction temperature?

From some sources (#1, #2), with images reposted below, there appears to be two ways to calculate the junction temperature.

Assuming the following thermal resistances, (a) junction to case = 0.5 deg. C/W, (b) junction to ambient = 20 deg. C/W, and (c) an ambient temperature of 27 deg. C.

Then, (1) the equation that uses the junction-to-case parameter results in a junction temperature of 42 deg. C.

(2) the equation that uses the junction-to-ambient parameter results in a junction temperature of 107 deg. C.

So what would be the appropriate interpretation of how to calculate the junction temperature of the transistor?

• Theta J-A often has assumptions about PCB layout (pad size, copper weight, etc) or heatsinking built in to the value. So if your conditions are different, the junction rise will be different. Theta J-C is not subject to those assumptions. Commented Apr 9 at 19:22

Junction to case. The junction to ambient thermal resistance is the junction to case thermal resistance plus the case to ambient thermal resistance, without any heat sinking applied (or with some specified heat sinking applied).

Since you know the case temperature, the junction to ambient thermal resistance is immaterial.