Thermal Resistance Testing of Power Modules: Standards, Methods, and Applications

Power module thermal resistance testing measures how efficiently heat generated at the semiconductor junction flows through the die-attach, substrate, baseplate, and thermal interface material to the case or heat sink, making it one of the most critical parameters for power density and long-term reliability. Using temperature-sensitive electrical parameters in accordance with JEDEC JESD51-14 and IEC 60747 standards, accredited third-party laboratories determine Rth(j-c), Rth(j-s), Rth(j-a), and transient thermal impedance Zth for IGBT, SiC MOSFET, and silicon power modules. This article explains the technical principle, sample requirements, test items, measurement procedures, and typical applications of thermal resistance testing for power modules.

Technical Overview

Thermal resistance (Rth) is defined as the temperature difference between the semiconductor junction and a specified reference point divided by the dissipated power, expressed in K/W. For a power module, the heat generated in the IGBT or MOSFET die must pass through the die-attach solder, the ceramic direct-bonded copper (DBC) or active metal-brazed (AMB) substrate, the substrate solder layer, the baseplate, and the thermal interface material (TIM) before reaching the heat sink. Every layer adds thermal resistance, and degradation of any interface — such as solder fatigue, voiding, or TIM dry-out — directly increases the junction temperature and shortens module lifetime.

Because the junction is physically inaccessible, thermal resistance is determined indirectly through temperature-sensitive electrical parameters (TSP), most commonly the forward voltage of the device body diode or the low-current on-state voltage of the chip. The measured cooling transient is converted into a transient thermal impedance curve Zth(t), from which steady-state Rth values and layer-by-layer structure functions are derived.

Key thermal metrics evaluated in power module thermal resistance testing include:

• Rth(j-c): steady-state junction-to-case thermal resistance
• Rth(j-s): junction-to-sink thermal resistance including the TIM layer
• Rth(j-a): junction-to-ambient thermal resistance under natural or forced convection
• Zth(t): transient thermal impedance as a function of time

The governing international standards include the JEDEC JESD51 series (in particular JESD51-1 for the electrical test method and JESD51-14 for the transient dual interface method), the IEC 60747 series covering IGBTs and isolated power semiconductor devices, MIL-STD-883 Method 1012.1 for legacy qualification, and the automotive qualification guideline AQG 324, which specifies extensive thermal characterization of power modules for traction inverters.

Testing Samples

Thermal resistance testing applies to virtually all power module families, from low-current discrete-based assemblies to high-power traction modules. Typical sample categories include:

IGBT Power Modules

Silicon IGBT modules in single-switch, half-bridge, chopper, dual, and six-pack (three-phase) configurations, covering both industrial-grade and automotive-grade products. The antiparallel diode or the gate-emitter on-state voltage at a small sense current is typically used as the temperature-sensitive parameter.

SiC MOSFET and Hybrid Modules

Silicon carbide MOSFET modules and hybrid IGBT/SiC designs, which operate at higher power densities and junction temperatures. Accurate Rth and Zth characterization is essential for these wide-bandgap packages because their thermal margins are often design-limiting.

Silicon MOSFET, Diode, and Thyristor Modules

Rectifier and thyristor modules for industrial rectification, welding, and soft-start applications, where junction-to-case thermal resistance determines the continuous current rating.

Intelligent Power Modules and Discrete Devices

Intelligent power modules (IPM) with integrated gate drive and protection functions, as well as discrete devices in TO-220, TO-247, and similar packages used for benchmarking or reference measurements.

Sample Requirements

Samples must be electrically functional with intact terminals and a baseplate free of warpage, scratches, or contamination that could affect contact with the cold plate. The device under test should be representative of production parts, and the reference point for the case temperature measurement must be clearly defined, typically on the baseplate directly beneath the active die.

Testing Items

Accredited third-party laboratories typically offer the following measurement items for power module thermal characterization:

• Steady-state junction-to-case thermal resistance Rth(j-c) per JEDEC JESD51-14
• Junction-to-heatsink thermal resistance Rth(j-s) with a defined TIM and mounting pressure
• Junction-to-ambient thermal resistance Rth(j-a) under natural or forced convection
• Transient thermal impedance curve Zth(t) from approximately 10 microseconds to steady state
• K-factor calibration of the temperature-sensitive parameter
• Die-attach, substrate solder, and TIM characterization via cumulative and differential structure functions
• Rth comparison before and after reliability stress such as power cycling, thermal cycling, and humidity exposure
• Batch-to-batch homogeneity and supplier benchmarking studies

The test report normally includes the calibration data, heating power, recorded temperature responses, the extracted Rth and Zth values, structure function plots, and a statement of measurement uncertainty.

Testing Methods

Temperature-Sensitive Parameter (TSP) Principle

The forward voltage of a semiconductor junction driven by a small, constant measurement current decreases almost linearly with increasing temperature. This relationship allows the electrical terminals of the module to act as an intrinsic temperature sensor for the die, eliminating the need to attach physical sensors to the junction.

K-Factor Calibration

Before power measurement, the sample is placed in a temperature-controlled chamber or on a precision hot plate. At several stable temperature points, the forward voltage is recorded at the chosen sense current, and the slope K (in mV/K) is obtained by linear regression. This calibration factor converts subsequent voltage transients into junction temperature transients.

Heating and Transient Capture

A defined heating current is passed through the device until thermal steady state is reached, establishing a known dissipated power P. The current is then switched rapidly, typically within microseconds, from the heating level to the small measurement level. A high-speed data acquisition system records the forward voltage during the cooling transient, which is converted into Tj(t) using the K-factor. The transient thermal impedance is calculated as Zth(t) = (Tj(t) − Tj0) / P, where Tj0 is the junction temperature at the moment of switching.

Transient Dual Interface Method

For junction-to-case resistance, the JEDEC JESD51-14 transient dual interface method (TDIM) is the industry-preferred approach. The module is measured twice on a temperature-controlled cold plate: once with a dry interface and once with a thin layer of thermal grease. Because the boundary condition at the case differs between the two runs, the Zth curves diverge at the point corresponding to the case. The thermal resistance value at the divergence point is taken as Rth(j-c). This method avoids the large uncertainties of legacy steady-state thermocouple measurements, where the case temperature reading is strongly influenced by sensor placement.

Cold Plate Control and Mounting Conditions

Reproducible results require tightly controlled boundary conditions: stabilized coolant temperature and flow rate, a defined quantity and spreading pattern of thermal grease, mounting torque according to the module datasheet, and accurate reference temperature sensors at the cold plate and case surface. Ambient temperature and humidity are documented for every run.

Structure Function Analysis

The measured cooling transient is mathematically deconvoluted into a Cauer-type RC network and rendered as cumulative and differential structure functions. These curves map the thermal capacitances and resistances of the physical stack — die, die-attach solder, substrate, substrate solder, baseplate, and TIM. Comparing the structure function of a sample against a known-good reference device reveals solder delamination, voiding, or TIM degradation, making the method a powerful quality and failure-analysis tool in addition to a datasheet measurement.

Applications

Thermal resistance testing supports the entire life cycle of power module products, from design to field failure analysis:

• Datasheet verification: confirming manufacturer Rth and Zth claims for catalog parts and customized designs
• Thermal design input: providing accurate models for heatsink selection, inverter thermal simulation, and derating curves
• Package and material development: evaluating sintered-silver die attach, AMB substrates, pin-fin baseplates, and TIM alternatives
• Reliability qualification: establishing baseline Rth before and after power cycling, thermal cycling, and mechanical stress sequences aligned with AQG 324 requirements
• Failure analysis: localizing solder fatigue, delamination, and bond-wire degradation through structure function shifts
• Application sectors: electric vehicle traction inverters, industrial motor drives, wind and photovoltaic converters, rail traction, and HVDC systems
• Incoming inspection and supplier benchmarking of modules destined for safety-critical installations

By quantifying how much power a module can dissipate before reaching its maximum junction temperature, thermal resistance data directly defines the continuous current capability, switching frequency limits, and expected service life of the final power electronic system.

FAQ

What is the thermal resistance of a power module?

Thermal resistance is the temperature rise per unit of dissipated power between the semiconductor junction and a reference point such as the case, heat sink, or ambient, expressed in K/W. A lower value indicates better heat extraction and higher permissible power dissipation.

Which standards govern power module thermal resistance testing?

The principal standards are JEDEC JESD51-1 for the electrical measurement method, JEDEC JESD51-14 for the transient dual interface method of determining Rth(j-c), the IEC 60747 series for the relevant device categories, and MIL-STD-883 Method 1012.1 for legacy qualification. Automotive modules are additionally characterized in line with the AQG 324 qualification guideline.

How is the junction temperature measured if no sensor can be placed on the die?

The measurement exploits a temperature-sensitive electrical parameter, typically the body-diode forward voltage or the low-current on-state voltage, which is first calibrated against temperature in a controlled environment. During the power test, the rapid voltage transient after switching off the heating current is converted into a junction temperature transient using this calibration.

What is the difference between Rth and Zth?

Rth is the steady-state thermal resistance reached after thermal equilibrium, while Zth(t) is the transient thermal impedance that describes how junction temperature evolves over time for short power pulses. Zth is essential for estimating junction temperature under pulsed or overload conditions.

Why is the transient dual interface method preferred for Rth(j-c)?

The traditional steady-state method requires an accurate case temperature measured by a thermocouple, which is highly sensitive to sensor position and introduces significant uncertainty. The transient dual interface method identifies the case plane from the divergence of two Zth curves measured with different interface conditions, giving reproducible results with markedly lower uncertainty.

What sample preparation is required?

The module baseplate must be clean and flat, thermal grease must be applied in a defined quantity and pattern, mounting screws must be tightened to the specified torque, and the sample must be electrically functional so that the temperature-sensitive junction can be driven and measured.

How long does a thermal resistance test take?

A complete characterization, including K-factor calibration, two or more cold-plate measurements, and structure function analysis, is typically completed within one to three working days per module configuration, depending on the number of switching positions and the required reporting depth.