The Demand for Seamless Characterization, Simulation and Development of Power SemiconductorsMarch 30, 2021 by Uwe Jansen
This article highlights publicly-funded project MessLeha addresses this issue by defining a machine-readable data sheet to support the setup of simulation models.
In recent years, the demand for power electronic systems has been steadily growing. At the same time, manufacturers have been faced with an increasing need to customize power semiconductor devices for specific applications. Currently, selecting the right semiconductor is challenging, as gaps in characterization standards impede the comparability of components, especially for fast power semiconductors.
Characterization methods and environments for semiconductors are defined in the IEC 60747 series  on semiconductor devices. However, the application of these standards is limited when it comes to new technologies like SiC and GaN. And even for the established technologies, the traceability of measurements is lacking.
Furthermore, setting up a simulation during the selection process is often laborious on the customer side. Semiconductor manufacturers have to decide which of the many available simulation tools they want to model. If their customers use a different tool from the ones they have selected, then parametrization will require considerable effort or the device won’t be considered.
The German, publicly-funded project MessLeha addresses this issue by defining a machine-readable data sheet to support the setup of simulation models. The project will also develop a measurement method and environment for measuring fast power semiconductors.
Figure 1: Overview of the project MessLeha
Development of a Digital Datasheet
In order to meet the requirements for a power converter, individual components must interact optimally in a complex system. For the power semiconductor, an initial selection is based on data sheet values. At a later stage, further criteria such as deliverability are added. Some power semiconductor manufacturers offer models for certain toolchains, or are in the process of preparing this step. At this point, the question arises whether this is the right way. Should a manufacturer of a component supply a model only for a specific toolchain? Are there other models to be created and maintained in the medium term? What about customers/users who have a different solution in use? What about market access for novel simulation tools if manufacturers offer models for one or maybe two specific toolchains? Does this perhaps even slow down technical innovation?
Neither semiconductor manufacturers, developers and manufacturers of power electronic systems nor manufacturers of simulation tools can have an interest in this scenario. It creates dependencies and generates unnecessary effort. The introduction of a machine-readable data sheet could be helpful at this point. The manufacturers of power semiconductors would store characteristic values as in the current PDF datasheet. Manufacturers of simulation tools would very quickly develop a solution for importing and automatically parametrizing component models, if they had such a coordinated data set. After the introduction of the machine-readable datasheet, developers would be able to evaluate the behavior of a specific component very quickly. Depending on the willingness of the manufacturers of power semiconductors to provide more data than usual in a current PDF datasheet, it would even be conceivable to provide reliability data or daily updated statements about deliverability. This would allow power electronics designers to further streamline the development process. A faster and more intense interaction between device vendors and their customers becomes conceivable.
It is highly likely that the aforementioned stakeholders will have a vested interest in the content of a machine-readable datasheet. The declared goal of the project is to explore this interest and bring their contribution together in an agreed standard.
Figure 2: Today (left), only customers with appropriate toolchains are able to use provided models, while the aim of future datasets (right) is to provide a defined interface suitable for all toolchains.
Development of a Modular Characterization Setup
Double Pulse Testing (DPT) is an important method for power device characterization and comparison. Characterization usually is done at the power device manufacturer’s to create data sheets and simulation models. Comparative device evaluation is performed by power device customers preparing strategic decisions between device and packaging technologies. For both cases, the devices need to be tested in a characterization setup that minimizes external impact on device performance as far as possible to attain consistent and comparable results. With the emergence of new Wide-Bandgap (WBG) semiconductors like SiC MOSFETs, link and current sensor, reducing parasitic capacitances, a fast and strong driver close to the device, and sufficient bandwidth for voltage measurement as well as for current measurement. For low-power applications, it is quite common to place the complete DPT setup including the DUT on a single PCB. However, at higher power, and for a larger product portfolio as common in the power module business, this approach tends to become cumbersome. Hence, a modular approach separating the complete setup into two main parts is advantageous. A device board, carrying only the device itself and an optional driver, is designed individually for the best fit of the specific module. As a counterpart, there is a DC link board containing a low-inductance DC link as well as a low-inductance current measurement. Of course, these two boards need to be connected by a unified low-inductance connection. Fig. 3 shows a proposal for such a board offering a simple low-inductance connection on the left and a DC link with several, optional current sensors on the right. This board is intended to house either a pulse current transformer, a planar shunt arrangement or a Rogowski coil.
Figure 3: DC-link board with low-inductance interface [email protected]
Comparison of Measurement Principles
Due to the fast switching times of Wide-Bandgap semiconductors, like SiC and GaN, the requirements of the equipment for the switching loss measurements are rapidly increasing in comparison to silicon devices.
The University of Stuttgart with its Institutes for Power Electronics and Electrical Drives (ILEA) and Robust Power Semiconductor Systems (ILH) is therefore working on improving and characterizing measurement setups for Wide-Bandgap switching loss determination. For this purpose, the existing state-of-the-art Double Pulse Test has been improved by new current sensors and high-accuracy characterization of current and voltage probes with regard to frequency behavior, i.e. bandwidth. Furthermore, the influence of the parasitics in the setup, like stray inductance, will be taken into account. The results are verified with the help of highly accurate calorimetric measurements, utilizing the heat-up phase for fast switching loss results at several points of operation.
In parallel, the National Metrology Institute of Germany (PTB) is developing a method for measuring the switching losses with a sampling measuring system. With this method, the voltage and the current will be recorded precisely during the switching time. For this purpose, the voltage divider and the shunt first need to be characterized. Another challenge is the time correction between the two recorded signals. These two factors ensure the accuracy of the calculated switching losses.
Figure 4: Power module as DUT and the different facets of switching loss characterization in this project (middle). Novel current sensor based on the HOKA principle for precise current measurement  (top left). Calorimetric measurement setup for Wide-Bandgap discrete power semiconductors  (top right). Bandwidth characterization for a shunt with a transmission line pulse generator (PTB) (bottom left). Precise modelling and error calculation of calorimetric and electrical measurements for level of trust calculation  (bottom left).
After the development phase of the three methods, a comparison will be performed between all methods. The measurement uncertainty of the systems will also be determined.
Standardization in IEC – and How to Participate
The project MessLeha aims to develop a standardized measurement environment and machine-readable data sheet to support a smoother development of power electronic systems. This can only be achieved if the solution is extensively updated. At the end of the project in December 2021, two drafts will be proposed to DKE, the German Commission for Electrical, Electronic & Information Technologies in DIN and VDE. Upon approval, these drafts will then be proposed to the Technical Committee TC 47 “Semiconductor devices” of the International Electrotechnical Commission IEC.
A proposal for an amendment to the IEC 60747 series will address the current lack of traceability of measurements. It will furthermore make sure that the Double Pulse Test will be applicable to the new SiC and GaN semiconductors as well. The second proposal will define the machine-readable data sheet.
On March 4, 2021, the project MessLeha will hold an online workshop to collect feedback from stakeholders. During two dedicated breakout sessions, the project partners will discuss with the participants their requirements for device model parametrization and simulations as well as for the measurement environment.
The input gathered during these sessions will be included in the ongoing work and the resulting standardization proposals. If you are interested in participating, you can learn more about the workshop content and registration on the website provided at the end of this article.
 Relevant are the following standards of the IEC 60747 series: IEC 60747-8 Semiconductor devices - Discrete devices - Part 8: Field-effect transistors IEC 60747-9 Semiconductor devices - Part 9: Discrete devices - Insulated-gate bipolar transistors (IGBTs) IEC 60747-15 Semiconductor devices - Discrete devices - Part 15: Isolated power semiconductor devices
 P. Ziegler, N. Tröster, D. Schmidt, J. Ruthardt, M. Fischer and J. Roth-Stielow, “Wide Bandwidth Current Sensor for Commutation Current Measurement in Fast Switching Power Electronics” in EPE'20 ECCE Europe: 22nd European Conference on Power Electronics and Applications, Lyon, 2020
 J. Weimer and I. Kallfass, "Soft-Switching Losses in GaN and SiC Power Transistors Based on New Calorimetric Measurements," 2019 31st International Symposium on Power Semiconductor Devices and ICs (ISPSD), Shanghai, China, 2019, pp. 455-458, doi: 10.1109/ISPSD.2019.8757650.
 D. Koch, S. Araujo and I. Kallfass, "Accuracy Analysis of Calorimetric Loss Measurement for Benchmarking Wide Bandgap Power Transistors under Soft-Switching Operation," 2019 IEEE Workshop on Wide Bandgap Power Devices and Applications in Asia (WiPDA Asia), Taipei, Taiwan, 2019, pp. 1-6, doi: 10.1109/ WiPDAAsia.2019.8760332.
About the Author
Uwe Jansen, Infineon Technologies AG,
Anna-Lena Heller, Kevin Hermanns, PE-Systems GmbH,
Florian Schilling, Ying Su, Physikalisch-Technische Bundesanstalt,
Dominik Koch, Julian Weimer, Philipp Ziegler, University of Stuttgart,
Alexandra Fabricius, DKE German Commission for Electrical,
Electronic & Information Technologies of DIN and VDE