This article presents the improvement of development-accompanying EMC measurement procedures for automotive applications in the area of interference emission. Currently, these tests are usually carried out with measuring equipment intended for component testing of vehicles. However, specially adapted EMC measurement methods can perform this task more quickly and cost-effectively. A practical example of such a measurement method is presented.

First, the peculiarities of the current measurement methods and the traditional approach to interference suppression are analyzed. Every assembly that is to be installed in a vehicle must be approved after the EMC component test. This test is first carried out with a sample assembly. If the module does not pass the test, the EMC engineer must rework it. Specified measuring stations such as antenna measurement or stripline measurement are used to determine the interference emission. The module is mounted on the measuring station with the cable harness.

With the antenna measurement method, the module and cable harness rest on a table. The cable harness is aligned to the antenna at a distance of one meter. The cable harness will typically be the source of the emission that can be measured with the antenna. The module itself is usually too small for its emissions to reach the antenna and be measured there. Near fields may be generated on the module by microcontroller operation, for example, but their intensity at the antenna is insignificant. The antenna only measures the module's near fields indirectly, namely as radiated emissions from the cable harness. Both internal RF currents and voltages as well as near fields may stimulate the cable harness to emit near fields. The situation is very similar in stripeline measurements. The cable harness is positioned under the stripeline conductor and couples RF into it. Not all of the module's near fields will couple to the stripeline conductor, especially if the module is beside the stripeline during the test.

The listed measuring methods and their characteristics show that they are not particularly suitable for a root cause analysis of a module's interference emissions. This means that these measuring methods can be used to assess RF emitted by the cable harness, but not to assess a module's potential near field coupling to its environment in a vehicle. The significance of near-field coupling in a vehicle will be demonstrated by taking the passenger compartment electronics (Figure 1) as an example:

Figure 1: Passenger compartment electronics with near-field coupling to the cable

The module is located at the front of the passenger compartment directly under the roof lining. The magnetic field of the microcontroller encircles the passenger compartment lighting cable and induces a voltage in this. This voltage stimulates the cable to function as a transmission antenna. The resulting radiated emission may interfere with sensitive vehicle components.

In practice, developers also use the antenna or stripeline measuring methods to optimize modules in terms of EMC. These measuring methods, however, are hardly suited to achieve a satisfactory optimization. The antenna or stripeline is unable to measure a module's near fields at the level that is required for optimization. The near field of the microcontroller (Figure 2) does not even reach the antenna. It is not detected by the measurement but may nevertheless cause interference in the vehicle later on (Figure 1).

This, however, has two decisive disadvantages:

1. The module's near fields, whose effect is visible in Figure 1, are not sufficiently analyzed (Figure 2).
2. The use of this measuring method during development is cumbersome, costly and time-consuming.

Figure 2: Measurement of radiated emissions from passenger compartment electronics with an antenna

The drawbacks mentioned under point 2 are due to the following:

• The device under test is connected to a cable harness. It has to be disconnected from the cable harness and taken to the developer's workplace for modifications that are performed outside the cabin. This takes a lot of time.
• The measurement setup often has to be reproduced for further measurements with an antenna or stripeline. But in most cases, the module and cable harness cannot be returned to an absolutely identical position. This results in measurement deviations.
• The developer has no direct access to the device under test during the measurement process since this is located in the closed cabin. The developer cannot carry out direct changes in terms of EMC optimization. Not even minor modifications are thus possible to improve the test result without great inconvenience. The setup is very inflexible when it comes to manipulating the device under test.
• The frequency response characteristics measured and the modification protocol cannot be compared immediately and flexibly. Here again, complex comparisons take a lot of time.

This shows that conventional measuring methods are inadequate. Efficient and productive EMC measuring methods are needed during development. All disturbances, particularly those that are effective in practice (Fehler! Es wurde kein Textmarkenname vergeben.), should be able to be measured flexibly and, if possible, directly at the developer's workplace.

Figure 3: Functional principle of emission measurements with an RF current transformer during development

Figure 4: Functional principle of near-field measurements with near-field probes in emission measurements during development

The diagram in Figure 3 show what has to be done with the measurement setup and which requirements this has to meet: 

• The developer must be able to measure the cable's RF current and trace his measurement directly with a spectrum analyzer (Figure 3). The measurement is carried out with an RF current transformer that is short-circuited to the ground. Wires that are interference-free or not involved should be disconnected or their resistance increased with ferrites. Ideally, the cable harness is limited to the power supply.
• The developer can detect the cause of the interference on the module with a near-field probe. The near-field probe measurement (Figure 4) must be able to be traced directly with a spectrum analyzer.
• Both measuring methods must be able to be recorded flexibly and allow the developer to compare measurements of both the same, though also different types.

Figure 5 shows a test bench that is suitable for the workplace of a developer or EMC engineer in the development stage. The shielding cabin here is implemented in the form of a shielding tent and can be placed on its conductive groundplane to shield the measurement setup from external electromagnetic fields. Power supplies and signals are led to the outside through the groundplane via filters. The front of the shielding tent can be folded up and down slightly. The entire shielding tent can be opened wide to allow easier modification of the device under test (Figure 5). The near-field probes can be connected to the spectrum analyzer through a shielded bushing in the groundplane.

Figure 5: Practical measurement setup of a RF current transformer with an electronic module (ESA1 test bench). The RF current transformer supplies the electronic module with power in this example.

The module to be tested is connected to the current transformer via a reduced cable harness. The device under test can remain in the shielding tent or is simply disconnected to carry out modifications.

The module's environment in the vehicle can be simulated with corresponding parts in the shielding tent. As far as the passenger compartment electronics (Figure 1) are concerned, the developer can simulate the relevant section of the cable with a corresponding tube. (Figure5) A current transformer (HFW21) or a line-impedance stabilization network (NNB21) is used to assess the induced voltage.

The frequency response characteristics that are measured are documented with a PC and customized software (Figure 6). This software allows the developer to record, color, annotate, calculate and visualize any number of curves of a spectrum analyzer. This enables a flexible, easy and fast comparison of the different steps of the measurement process. The developer can simply export images and data from the software for documentation and statistical analysis.

Figure 6: Remote control of the spectrum analyzer and documentation of the results with ChipScan ESA

It is important for developers and EMC engineers to be able to find the causes of interference on modules and also test the effect of modifications immediately at their workplace. This results in noticeably lower costs and less time for the development of modules and devices.