Success Factors of EMC in the Design PhaseMay 28, 2020 by Jan Spindler
This article discusses electromagnetic compatibility and how it can prove difficult for developers and project managers.
Electromagnetic compatibility (EMC) can prove difficult for developers and project managers. The first EMC testing results are rarely as good as expected, and additional obstacles may arise during development. These unexpected changes can result in shifted schedules, increased development costs, and external post-testing costs.
EMC At A Glance
EMC is the ability for electrical equipment to function properly in its electromagnetic environment without introducing intolerable disturbances to other equipment. However, few topics are as difficult to grasp as EMC. You cannot see, hear, or feel electromagnetic waves, which makes it difficult to determine influencing factors and structures.
Engineers today face roadblocks of greater complexity, as smaller package sizes often conflict with the economic goal of reducing cost and development time. Because of this, compromises must often be made. This intense pressure to deliver quickly often leads to EMC issues that only come to light during a project’s final phase. MPS’s goal is to offer technical support to customers and prepare them for future technical challenges. With a clearly-structured EMC development process, these issues are easily avoided (see Figure 1).
Figure 1: Structured EMC Development Process
Requirements and Testing Plan
Begin with the project schedule. A schedule ensures everyone involved knows the goals and requirements to sell a product to the end customer. Depending on the product and industry, different requirements must be fulfilled. On top of various regulations, international standards and national standards must be met (see Figure 2).
An example of this is the automobile industry, where the complexity of vehicle development involves additional requirements that ensures appropriate component interactions. The automobile industry spent years defining the requirements for electronic components and subassemblies. Limits and test levels have become standardized, and good and fast risk assessments can be made based on component measurements.
Figure 2: Example of Commercial Standardization
However, meeting the legal requirements is not enough. Nearly every car manufacturer has their own EMC requirements with higher test levels and lower emission limits. Strict compliance with these requirements ensures that communication services (broadcast and mobile) and high-power electronics (e.g. e-boosters, e-trains, onboard chargers) work in parallel.
This planning process also benefits other industries, where products consist of subassemblies — some of which are not even developed in-house. While there is no guarantee that all limits and test levels will be met when the module is interconnected, defining requirements in advance significantly reduces this risk.
A unique testing plan should be integrated into the development schedule. This plan is a working document to be updated during the first phase of the development and acting as a central document for the testing and documentation phase of the project.
A good test plan defines and documents all relevant information about the equipment under test (EUT). This document should allow an engineer to run their tests without additional documents and background knowledge. Defining operation modes during testing is as important as the test set-up (grounding, wiring harness, load box, and software configurations). Companies who work with external suppliers and multiple customers should document all mutually understood requirements in the test plan. Requirements should be tracked throughout the process, especially for purchased parts.
Concept and Technology
PCB size and connector positions for the cables are usually predefined by mechanical limitations. The housing is based on cost structure, weight, and thermal concept.
Mechanical designs require great attention be paid to EMC. Decisions made in this stage can result in later problems that can only be solved through extensive efforts in redesigning the layout or conducting additional EMC measurements.
For example, the initial positions of plugs and function groups on the board (e.g. the switching regulator) are important, as the switching creates E and H fields. The direct field coupling is determined by the distance, design, and heights of the components (e.g. the main coil) (see Figure 3).
Packaging, efficiency, and costs drive the development of switching regulators. Products are being delivered with higher switching frequencies, including ICs with a frequency-spread spectrum and highly integrated power modules.
Consider the advantages and disadvantages of a higher or lower switching frequency when creating your design. Automotive suppliers often comply with limited AM band values at switching frequencies below 1MHz. The fundamental frequency, or the first harmonic of the fundamental frequency, lies within this band. In addition, relatively large input filters and large coils are sometimes necessary at low switching frequencies.
A switching frequency above the AM band (e.g. 2MHz) offers the advantage of the fundamental frequency and the first harmonics being outside of this band. Components such as the input filter and coil can be reduced by size, value, and cost. However, there are disadvantages like converging on switching losses, thermal rise, and EMC. Remember these tradeoffs when determining which topology and technology to use.
Table 1: Choosing the Switching Frequency (2)
Figure 3: Field-Coupling on the PCB (1)
Layout and Review
At this point, conduct a risk assessment based on empirical values, the analysis of possible technologies, and the evaluation of purchased parts and suppliers. Begin working on potential sources of interference and the product’s weak points, since the layout consists of compromises (mechanical, thermal, functional, and EMC-technical). Before the first board is produced, an internal or external layout review helps to find and implement the best compromise.
In the first layout, additional pads and components can allow you to try different tactics during EMC measurement. You could also create variations of boards and assembly options during the first measurements at an EMC lab, since several simultaneous tests will allow you to gather results quickly, such as information regarding complex coupling. No matter how you plan in advance, optimal results are difficult to define without testing.
Testing and Optimization
Before testing your system, consider the test results from possible assemblies and components. These previous steps dealt intensively with the product itself, and they offer a good overview that will help you determine which topics to focus on in the pretests. With every test, the project becomes more transparent, and project risk is reduced.
The first tests with a real-world application should take place next. Pretests are key to finding weak spots early on. Based on the test plan and experiences with similar products, filter out individual test methods and check them in advance. It is recommended to conduct a combination of interference emission measurements and immunity measurements. A solid overview and quick feedback is important. The test plan is particularly helpful in preparing the EMC measurement for coordination with an EMC lab. There are two possible results to obtain from the EMC lab:
Result 1: EUT does not pass the requirements
In this case, the sources of interference must be identified, and optimization measurements should be obtained. If appropriate assembly options have been provided at critical points, and alternative components are included, you can start optimizing right away. Measurement time is limited. Even if you do not have your own equipment for an EMC analysis, gather as much information from the test session as possible.
Result 2: EUT passes the requirements
If the product fulfills the EMC requirements, you will have opportunities for optimization. The ideal scenario between costs and results can often only be determined by multiple measurements.
About the Author
Jan Spindler holds a Bachelor's Degree in Electrical Engineering at the Cooperative State University Stuttgart. He currently works as the Head of the EMC Laboratory at Monolithic Power Systems in Ottobrunn, Germany.
This article originally appeared in the Bodo’s Power Systems magazine.