Съвременни методи, използвани при проектирането на потребителска електроника

Imagine for a moment that you are testing a completely new car: you sit in one of the ergonomic seats, start the engine, and to complete the experience, you want to listen to music. However, when you turn on the radio, you suddenly notice some interesting side effects. For example, a light in the passenger cabin turns on. Or the radio starts to squeal and buzz. These are relatively harmless phenomena. However, much more serious problems can occur, which can even threaten driving safety or limit the functionality of the car. High-frequency interference is very troublesome in modern electric cars, hybrid vehicles, or complex electronic systems that ensure driver safety and comfort.

Fig. 1 The component radiates to the metal structure of the driver's seat via the electric near field. The driver's seat, excited in this way, emits waves to the vehicle's antenna, interfering with the radio signal reception.

Sources of emissions and limitations of traditional EMC measurements

Usually, it is not the entire electronic component, but individual devices that are responsible for emissions. Devices such as a quartz resonator or microcontroller create local electromagnetic fields that induce voltages in enclosures or structural parts. These parts are thereby stimulated to oscillate, and as a result, they emit radiation. Software developers will certainly check their components for emissions by performing component measurements in an EMC test chamber during development. However, these measurements will generally capture emissions from the tested device as a whole. The measurement methods used in such tests cannot sufficiently assess the near fields of the tested device. During conventional component measurements, the designer cannot directly interact with the tested device, measure individual sections of the component in detail, or take actions to reach the emission source. The tested device is removed from the EMC chamber, placed in packaging, and brought back to the workspace. Another problem with these measurement methods is that the component is tested outside the actual vehicle environment, and emission frequencies may not be detected because other nearby structural parts are not excited to oscillate.

Back in the workspace, the developer can only compare the frequency plots from the component test with those obtained from experience in other development processes and formulate hypotheses about the causes of problems. Then, they modify the tested device based on these hypotheses. Only further test measurements in the EMC chamber show whether the change was correct and effective. The measurement setup must be repeated for new component measurements. However, in most cases, the component, especially the wiring harness, cannot be returned to exactly the same position. This causes measurement deviations. Frequency response characteristics measured at different stages of development cannot be compared immediately and flexibly. The designer must place individual protocols side by side and compare them step by step. This approach to EMC component development is time-consuming, complicated, and unsatisfactory for the designer—it requires significant effort and incurs high costs.

The need for more effective interference source detection

In this case, more effective searching for emission sources in complex electronic systems during development is necessary. The engineer must be able to measure as many disturbances as possible, systematically identify RF sources, flexibly make modifications, and conduct tests in their workspace to save time and costs.

Let's now look at how a component can become an emission source. The electronic assembly or PCB itself usually does not emit any waves. However, individual devices can generate RF near fields that surround, for example, connected cables. These induce voltage in the cables, causing them to emit waves. Due to electrical or magnetic coupling (i.e., in the near field), the entire metal system, including the component and connected cables, as well as metal parts such as enclosures, shielding plates, etc., in its immediate vicinity, is subject to self-excitation. (Fig. 2)

Fig. 2: A microcontroller placed on a component emits a magnetic field. It surrounds the steering column, where it induces a voltage. This voltage stimulates the steering column to emit radiation, which can interfere with sensitive components near the driver's seat.

Modern methods for emission measurement in the engineer's workspace

The entire metal system acts as an antenna when excited by electronics. The RF drive current flowing from the electronics to the antenna (cables and metal parts) can thus be considered an approximate measure of emissions from the tested device. We will now describe a measurement setup that meets our requirements for effective emission development of components. Emissions are measured under conventional measurement conditions, for example using an antenna. The device under test must be modified if one or more frequencies of the development sample exceed the limit values specified in the relevant standard. Values obtained from this measurement are used as a reference point for subsequent comparative measurements. (Fig. 3)

Fig. 3: ESA1 system with ChipScan-ESA software and a spectrum analyzer.

Near-field measurement and the use of a shielding tent

When measuring emissions from electronic circuits in the developer's workspace, it is important to define a measurement setup that simulates the component's environment in the passenger cabin as accurately as possible. A closed measurement configuration is set up in the workspace to measure all reasonable RF currents from the tested device. If the relevant frequencies fall within a range strongly affected by external disturbances, such as radio frequencies, a shielding tent is used to protect the measurement setup from these interferences. The shielding tent has a surface of 50×90 cm and attenuation of over 40 dB in the frequency range from approximately 80 to 650 MHz. The front part of the shielding tent can be folded up and down. The entire shielding tent can be widely opened to facilitate modification of the tested device. Power cables, such as electrical supply and measurement signal cables, are routed outside through filtered feedthroughs in the ground plane. This also creates a fixed grounding for the tested device and measurement equipment.

How are measurements carried out in the developer's workspace? In the first stage, comparative measurements are performed using RF current transformers inside the closed shielding tent and documented. The results of these measurements are compared with the component measurement results to confirm the accuracy of the measurement setup. Of course, the measurement results will not match. However, it is important that the relevant frequencies from the component measurement also appear in the frequency plot of the selected measurement setup.

Identification and elimination of RF radiation sources

Next, magnetic and electric field probes are used to scan integrated circuits, linear connections, plug connectors, etc. Precise frequency analysis and near-field orientation often allow the developer to recognize correlations between fields and driving currents. To measure near fields using probes, the front of the shielding tent must be opened. Of course, the shielding effect is then much lower than in the closed tent. However, the near fields of the tested device are usually much stronger than fields coupled from the environment, so measurements can still be successfully carried out. Field intensity at a specific frequency and within a frequency range (Fig. 4 and Fig. 5).

Fig. 4

Fig. 5: Use of electric and magnetic field probes for near-field measurements on components and their devices.

Possible sources of radio radiation include:

• electric fields above devices such as processors,
• electric fields on switched lines and bus systems,
• magnetic fields on switched data and clock lines,
• magnetic fields in power supplies.

Once RF interference sources are identified, the board can be modified on-site by soldering components, applying shielding, or rerouting wires. The results of further measurements using the RF current transformer in the closed shielding tent immediately show whether the measures taken are effective or not. RF current transformers and near-field probes can be used alternately in subsequent tests. The element can be continuously modified until the minimum excitation current from the RF current transformer flowing in the power line is achieved. Measured frequency response characteristics can be documented using a PC and custom software. This software allows developers to record, color, label, calculate, and visualize any number of spectrum analyzer curves and enables flexible, easy, and quick comparison of different stages of the measurement process. Developers can simply export images and data from the software for documentation and statistical analysis. (Fig. 6)

Fig. 6: Large amounts of data provided by the spectrum analyzer can be conveniently measured and compared using ChipScan-ESA software.

Testing connectors and ICs for EMC

A closer look at component electromagnetic compatibility issues shows that plug connectors, especially those used in high-voltage systems of hybrid cars, are unique. They are subject to high EMC requirements. Testing such connectors on a component prototype is actually too late, as the designer cannot improve the connector. Therefore, the designer should test individual connectors for their EMC characteristics early in the design process or rely on connectors with known EMC parameters that are considered sufficient for a given component.

Fig. 7: The designer could identify the emission source in Fig. 1 using near-field detection. In this case, radio reception interference could be prevented by moving the emission-causing element to a location in the car where no parts can be stimulated to oscillate nearby.

In addition to devices with one or two cable connections, complex devices can also be tested. At the beginning of the test, various effects caused by different RF sources in the tested device overlap, leading to reinforcement or partial attenuation of RF fields at specific frequencies. Therefore, to conduct an effective root cause analysis, especially for complex tested devices consisting of multiple PCBs, it is important to disassemble the device into manageable parts and examine individual PCBs separately. The tested device shown in Fig. 8 has several potential emission sources. In this example, the examination is limited to the interface module of the component.

Fig. 8: Possible emission sources of the interface module in a complex system.

We can imagine three RF sources:

1. connector between the base assembly and the interface module

2. electronics (PHY with microprocessor) in the interface module

3. interface cable connector

These three RF sources will now be discussed sequentially. This requires a measurement setup that suppresses other RF sources and the base assembly RF sources as much as possible.

1. Connector between the base assembly and interface module (Fig. 9)

The base assembly and interface module are connected via data and control lines. These lines are well protected on the PCB in the base assembly and interface module area as they are embedded in grounded surfaces. However, in the connector area, they pass freely through open space. Signal currents in the lines generate RF magnetic fields that spread inside and around the connector. They induce voltages on the connector ground pins. As a result, there is a voltage difference between the base assembly and the interface module. This causes an RF current to flow into the interface cable via the interface board, which in turn excites the cable, causing emissions from it. The current transformer COM port is connected close to the GND of the interface module to measure the current driven by the induced voltage. The measured current serves as a measure of the connector's contribution to the total emissions from the tested device. The effect of modifications, such as filters on signal lines or pin reassignment, can be measured directly.

Fig. 9: First stage of measurement using ESA1 – detection of emission sources at the connector between the base assembly and interface module

2. Electronic circuits (PHY with microprocessor) on the interface module (Fig. 10)

ICs of the interface module generate currents flowing to the GND system, inducing voltage. This voltage drives current from the base assembly to the connected interface cable, which excites the interface cable to emit. The base assembly remains connected to the ground plane during measurement. The connector's contribution to emissions between the ground plane and interface module is offset by several large-area GND connections. The voltage induced by the IC currents can be captured using a current transformer located at the top of the assembly. Modifications performed directly on the interface module can be evaluated in this way.

Fig. 10: Second stage of measurement using ESA1 – detecting emission sources in electronics

3. Interface cable connector (Fig. 11)

The interface connector is another RF source in this complex component. During data transfer, the interface driver passes current through the connector. This current induces direct-axis voltage in the connector housing, stimulating emissions from the interface cable. The current transformer is connected to the interface cable to perform the measurement. Note that during this measurement, the voltage is applied according to point 2 above. This voltage can be shorted by attaching a piece of copper foil to the interface PCB.

Fig. 11: Third stage of measurement using ESA1 – detecting emission sources at the interface cable connector.

The IC is another electronic element that is important for the component designer. The characteristics of ICs used in the electronic system greatly influence the EMC behavior of the entire component. Microcontroller and chipset structures are becoming smaller. Therefore, IC sensitivity may now be up to ten times higher than previous models. The behavior of ICs and their package types in terms of immunity and emissions is a key aspect that the designer must consider when selecting devices for the component. Therefore, it is important to choose the appropriate IC and use it according to relevant EMC requirements during the component development planning stage. Compliance measurements according to BISS/IEC standards should be carried out on ICs with respect to EMC parameters as a standard procedure. However, these measurements are not sufficient to ensure that the IC will perform in practice. Moreover, ICs should be tested using practical and universal EMC parameters, such as ESD. Impulses occurring in ICs during ESD and pulse tests for devices can be simulated in IC immunity tests. Special IC measurement and testing technology should enable the designer to perform immunity tests independently of the device or component. (Fig. 12)

Fig. 12: Shows the test configuration for direct interference coupling to the LFBGA package. The IC is mounted on a special IC adapter board. Interference can be introduced directly to the BGA balls using P200/P300 series probes. External IC wiring complies with the manufacturer's specifications. Additionally, filter elements are used in the power and signal lines to prevent the interference pulse from discharging and thus ensure defined conditions.

During IC emission analysis, the designer must thoroughly examine the physical process of the component. For example, if a fast high-frequency current and voltage circuit is placed in an unfavorable position in the device, it can interfere with the operation of the component itself or other components or devices through coupling paths via the component.

Conclusions: Effective EMC planning in component development

1. Test results help achieve three goals:

2. Improving the IC to avoid problems during later use

3. Practical parameters and conditions for interference-free use of the IC in components

4. Selection of an IC suitable for the user's application based on IC EMC parameters. This gives the electronics developer access to procedures and information that help plan EMC for a complex component in advance and measure and modify it directly in their workspace during development. Development becomes more efficient and less time-consuming. Fewer component tests in the EMC chamber will be required. This reduces long waiting times for using the EMC chamber. This, in turn, accelerates component development, saving resources, time, and costs.