Confusion with relating thermal resistance temperature, and electrical power and heat sinking

Question

What is meant by the following?:

In general, a device with a thermal resistance θ equal to 100°C/W will exhibit a temperature differential of 100°C for a power dissipation of 1 W, as measured between two reference points.

I don't quite understand the relation between the dissipated power and temperature rise. What should be the way of thinking here?

The device datasheet has both max voltage current and power ratings; and it also has the max junction temperature given. How to relate those and what to do is a bit vague for me.

Answer 1

What should be the way of thinking here?

Due to the θ_JC the heat flux moving from the junction to the case raises the temperature. Increase either thermal resistance or heat flow and temperature rises.

A low θ_JC is better than low.

There are multiple thermal resistances that may be specified. The most common are

• θ_JC
• θ_JA

θ_JC is the thermal resistance from the heat generating junction(s) to the case or thermal solder pad of the device package.

θ_JA is the thermal resistance from the heat generating junction(s) to ambient air. This includes the PCB and or heatsink.

θ_JC is a function of the geometric design of the device and is fixed with minor variation due to manufacturing.

θ_JA is a reference for comparative purposes. Because θ_JA includes the PCB and heatsink design, this parameter is measured using standardized test conditions typically specified by JEDEC in the EIA/JESD51 documents.

Thermal design terminology is explained in this document: Semiconductor and IC Package Thermal Metrics

θ_JC is what you use to estimate the junction temperature (T_J) of the device. T_J is mostly a factor of PCB design and secondly the device package design which is represented by θ_JC. Power dissipation is a very minor factor.
Source: Semiconductor and IC Package Thermal Metrics

This Simplified Thermal Resistance Model from Thermal Design By Insight, Not Hindsight shows the various parameters of interest in PCB thermal design. It also gives the basic calculations to consider for thermal design.

---

Other references:

The Effect of PCB Design on the Thermal Performance of SIMPLE SWITCHER Power Modules

Understanding Thermal Dissipation and Design of a Heatsink

---

Answer 2

You have power being dissipated in the junction, that energy flows away thru a thermal resistance which develops a temperature difference between the junction and whatever the other reference point is proportional to the power flow times the thermal resistance, it is just the thermodynamic version of ohms law.

You must contrive to respect all of the maximums, and that usually means that you will not be able to for example run max voltage and max current at the same time without fairly heroic efforts at cooling.

If say the thermal resistance between junction and ambient is 100C/W, and you have a value for Tj(max) of say 175C, then if the ambient temperature is limited to say 40C, you have a thermally imposed power limit of (175 - 40)/100 = 1.35W.

I would further note that you don't generally want to be anywhere near the maximum junction temperature as that hurts reliability.

Be very careful about the numbers on the first page of the datasheet, these are usually written by marketing slime rather then engineers, and while not wrong exactly usually do things like giving mosfet ratings at 25C case temperature, and Rds(on) ratings at 25C junction temperature, yes, mosfet vendors looking at YOU. The real usually starts on page 2 or so.