A modular and scalable system for electromagnetic compatibility testing of integrated circuits

In modern vehicles, electronic components are responsible for an increasing number of functions. From keyless entry systems to airbag deployment and anti-lock braking system. Electronic components are responsible for a safer and more pleasant driving experience and are nowadays installed in a wide range of electronic control modules in the smallest of installation spaces. However, with so many electronic control modules containing numerous ICs in close proximity often operating at ever-faster switching rates, the electromagnetic environment in the car poses significant challenges to the EMC for these components. Each IC emits electromagnetic emissions when active, which can interfere with other electronic components, resulting in malfunction or failure.

Especially for automotive ICs, the generic IC EMC Test Specification [11] specifies a series of EMC tests that an IC must pass to be used in a vehicle. Such tests range from conducted electromagnetic emission (EME) tests and radiated EME tests to electromagnetic immunity tests and transient tests such as ESD. This specification covers a wide range of characterizations necessary to release an IC for the automotive industry. At the moment, IC manufacturers usually build a special EMC test board for each of these tests, which must meet all the very specific requirements of the standards and the Generic IC EMC Test Specification. These test boards are then specially designed for the specified frequency range. For example, the frequency range for conducted EMC tests is \(150\,\text{k}\text{Hz}\) to \(1\,\text{G}\text{Hz}\). Designing a functional test board that meets all the requirements is already a not very easy and usually cost-intensive challenge, especially if it is to be designed for high load currents of the ICs. The construction of such test systems for each new product alone is a laborious and costly process with the traditional method. Since the trend in microelectronics is towards ever faster switching cycles and higher data transmission rates in communication systems, characterizing ICs just up to \(1\,\text{G}\text{Hz}\) will no longer be sufficient in the future. Therefore, it is becoming increasingly important to evaluate the electromagnetic emission and immunity of an IC also at higher frequencies in order to make statements about whether the IC itself or possible other ICs in the system are disturbed. There are already a number of publications describing an improvement and frequency range extension for the so-called DPI method, which is often used to characterize the electromagnetic immunity of ICs. A corresponding design approach with which the usable frequency range could be extended up to \(2\,\text{G}\text{Hz}\) was explained in [9] and [2], for example. An optimization of the DPI test bench for highly reflective low voltage devices was presented in [7]. A further proposal to extend the frequency range of the DPI method even further (up to \(20\,\text{G}\text{Hz}\)) is presented in [8]. The power transmitted to an IC in a SOIC8 package is thereby determined by offline calibration using short-open-load-thru (SOLT) or thru-reflect-line (TRL) using a corresponding printed circuit board (PCB) design. Another possibility of coupling interference signals into an IC by means of an off-board probe is presented in [1]. In this paper, the DPI investigation of a low-dropout regulator (LDO) is used to compare the conventional DPI method with direct RF power injection using a probe. The experimental results show a maximum deviation of the immunity level less than \(1\,\text{dB}\).

Another challenge with DPI testing is that during the test, the device under test (DUT) can be broken by the injected power level or a malfunction caused by it. In this case, the IC must be removed from the PCB. Due to the soldering and desoldering process, the PCB is often so damaged after a certain number of changes of the DUT that testing is no longer possible. In this case, the entire test board, including all components, must be replaced and renewed.

At present, all semiconductor manufacturers produce their own test boards. Although the standards specify what a test board should look like, it is up to the manufacturers themselves to ensure that the standards are met. On the other hand, at the system level, there are many individual components, such as coupling decoupling networks (CDNs), current clamps, etc., which are tested and certified and can also be used in combination according to individual specifications. Nothing like that exists at IC level yet. This reduces the comparability of the results as the measurements are highly dependent on the design of the test board.

To address these issues, a new modular approach has been developed to construct a cost-efficient and reusable test system that also allows the coupling of interfering signals over a higher frequency range. Such a modular system has already been developed and validated for ESD tests [10]. This system makes it possible to replace complex ESD evaluation boards with a kit of different modules in a more cost-effective, flexible, and reusable way.

This approach involves outsourcing various standard modules required for each test board to their PCBs and utilizing a separate PCB for the device under test.

By assembling the required measurement circuit using a kit principle and connecting the appropriate modules, the test board can be easily customized to meet the specific needs of the IC being tested. This approach not only reduces waste and costs but also allows for a more flexible and scalable testing process.

This paper is structured as follows: In Sect. 2 we describe a commonly used method to characterize the electromagnetic immunity and the electromagnetic emission of an IC where the modular system can be used, the DPI and the \(150\,\mathrm{\Omega}\) method. In Sect. 3 we briefly introduce the DUT, which is used as an example to demonstrate the modular test system. In Sect. 4 we give an overview of the key parameters of the modular system. In Sect. 5 we explain the different types of modules, which can be used to create the test system. Sects. 6 and 7 show examples of test setups for EMC characterization built with the presented modules and finally, in Sect. 8 we conclude the findings and give a short outlook.

The most important standards for electromagnetic compatibility testing at IC level are the IEC 62132 standard for electromagnetic immunity and the IEC 61967 standard for electromagnetic emission characterization. Two of the most important measurement methods for conducted immunity and emission tests are described in this chapter.

The modular test system is demonstrated for an automotive smart power high-side switch. A block diagram of this IC can be seen in Fig. 1. Automotive smart power high-side switches are commonly used in vehicles to switch electrical loads such as headlights. The integrated DMOS power transistor performs the switch function with a minimal on-resistance of \(R_{\text{ON}}\approx 4\,\text{m}\mathrm{\Omega}\) and a nominal load current of \(20\,\text{A}\). In addition to the DMOS, the IC has a number of protective functions such as overtemperature and overload protection as well as several diagnostic functions including current measurement and indication of failure states such as overheating. The IC’s diagnose output provides current measurement during rated operation and an indication of fault conditions. Because of these additional functions, the power switch is considered a smart power switch. Such switches are usually located in the vehicle’s control unit and are connected by a wiring harness to the battery and the switched load. The cable harnesses can act as antennas for electromagnetic interference that may interfere with the operation of the IC.

This section provides an overview of the modular test system by presenting its general parameters. The test system comprises a base module, which is a large metal plate including numerous screw holes used to secure and interconnect different modules using screws. On top of the metal plate, a gold plated PCB topper is placed, to ensure a reliable grounding of the screwed-on modules [10]. The modules are classified into standard modules, such as coupling modules, CDN modules, inductor modules, and the DUT module. The DUT module includes all the necessary components for operating the DUT.

One of the key benefits of the modular approach is its reusability. Standard modules can be employed for future tests, leading to cost savings. Separating the modules into DUT and standard modules is also cost-effective. In conventional boards, when the DUT breaks during immunity tests, the only option is to desolder and replace it, which can be done only a few times before the entire PCB with all components must be replaced. In contrast, the modular system only requires replacing a broken DUT module, allowing the use of higher-quality components in the standard modules, that can extend the frequency range of the test system and conduct tests in the higher frequency range in which many of today’s disturbance signals are present.

The modules are manufactured using a 4-layer FR4 PCB. The first layer contains all RF and signal lines. The impedance-controlled RF traces are designed as grounded coplanar waveguides and allow a \(50\,\mathrm{\Omega}\) system from the interference source to the IC pin. This impedance-controlled system minimizes reflections and losses in the coupling path and increases the possible frequency range in which the test system can be used. The second layer is a solid ground layer that serves as a current return path for the RF currents. Layer three is used for signal traces that cannot be routed on the top layer, while the bottom layer is a blank solid ground plane. By screwing the modules onto the large metal plate, the bottom plates of the modules are pressed against the large metal plate, establishing a solid ground connection. The PCB stackup including the mounting plate, where the modules are screwed on is illustrated in Fig. 2.

To exhibit the system’s modularity, Fig. 3 shows a demonstrative circuit that concurrently merges three distinct EMC test setups. T1: DPI for local pins connected to a component on the same PCB. Additionally, DPI for global pins connected to a cable harness is measured at T2. The \(150\,\mathrm{\Omega}\) emission measurement is indicated as T3. The individual modules in the setup are discussed in the following section.

The testing system comprises numerous standard modules designed for various testing methods. In this section we introduce all the standard modules, it takes to build two standard IC level EMC tests. The DPI setup for immunity tests and the \(150\,\mathrm{\Omega}\) setup for emission measurements. In the Sects. 6 and 7, these modules are combined to the mentioned test setups and evaluated.

A novel modular approach of a test system for the conducted EMC characterization of ICs was introduced. Conducted EMC tests were performed on an automotive smart power high side switch in order to investigate the advantages, such as the modularity and the increased frequency range of the proposed system. Using the different modules, a DPI testing setup and a \(150\,\mathrm{\Omega}\) measurement setup were assembled. In conclusion, the use of a modular test board for immunity and emission testing offers several advantages over the traditional approach of designing an EMC test board such as reduced design efforts, costs, and a significant increase in the injected signals frequency. Implementing standardized modules could potentially reduce the impact of test board designs on the measurement results. As with system-level testing, where standardized and certified components are utilized to ensure comparative and accurate measurement, standardized and certified modules could be implemented at the IC level to enhance the comparability of individual measurements and facilitate more straightforward test setup design for manufacturers, furthermore, the design costs and waste can be reduced because of the reusable design of the modular system. The system can easily be extended by other modules which are needed for other test setups. The next modules that are planned are symmetrical coupling modules which are needed for characterizing CAN bus IC pins.