As an electromagnetic compatibility (EMC) engineer who has spent years working on vehicle-level test development, I would like to share some practical experience from my research on intelligent electric cars. The rapid growth of vehicle electrification, intelligent driving, and telematics has greatly changed the electromagnetic environment around a car. The failure mechanisms that occur in a traditional internal-combustion vehicle are no longer sufficient to describe the risks found in an electric car. In particular, a modern intelligent electric car is equipped with dedicated short-range communication components, on-board telematics control units, high-voltage charging systems, and portable charging equipment. Each of these subsystems introduces a unique set of EMC challenges. When I was asked to establish a complete EMC validation plan for our electric car development projects, I realized that the existing standards do not always provide detailed guidance for the particular test scenarios that we face every day in the laboratory. In this article I would like to summarize several important test cases, the standard requirements behind them, and the practical methods that I have used to solve real-life problems during the development of intelligent electric cars.
Introduction to Vehicle-Level EMC Work on Intelligent Electric Cars
The modern vehicle is no longer a simple transport machine. It has evolved into a connected node that participates in the intelligent transportation system and the broader Internet of Things. The most visible innovations on an intelligent electric car include the electronic toll collection (ETC) system, the telematics box often called the on-board communication unit (OCU), and the various charging components that allow the driver to recharge the high-voltage battery either from an alternating-current supply, a direct-current fast charger, or a portable charging cable. All these systems are intended to make the electric car more convenient, more efficient, and safer. However, they also create new paths by which electromagnetic energy can be coupled into sensitive circuits, degrade radio reception, disturb the communication link between the vehicle and the roadside equipment, or even affect the safety-related functions of the charging process.
From my viewpoint as an EMC practitioner, the development of an intelligent electric car is a multidisciplinary task. The high-voltage traction system may generate broadband switching noise. The digital processors in the telematics unit can produce clock harmonics in the frequency ranges allocated to mobile communication. The ETC unit has a transmitter and receiver operating in the 5.8 GHz band and may be affected by radiated fields coming from other vehicle systems. The charging process involves a connection between the electric car and an external power grid, so conducted emission and immunity requirements become particularly important. A significant part of my work has therefore been focused on determining the correct test setup, identifying the most appropriate monitoring parameters, and designing practical fixtures that allow us to judge whether a device under test is functioning correctly under electromagnetic stress.
In this article I do not intend to present a theoretical review of EMC. Instead, I want to describe the methods that I have used in my own laboratory work, the criteria that I have applied to evaluate the results, the complications that I have encountered, and the ways in which I have solved them. The three main sections are devoted to ETC, charging equipment, and telematics control units. For each topic I will first discuss the applicable standard framework. Then I will describe a test arrangement that has proven useful in vehicle-level testing, the evaluation criteria that were selected, and the results obtained from sample vehicles. The last part highlights the benefits of these practical methods and discusses the direction for further development.
EMC Standard System for Intelligent Electric Cars
Before starting any test campaign, I always examine the standard system that applies to the vehicle or component under development. The EMC testing of motor vehicles is becoming more and more complex because the type-approval framework now includes not only classical radio disturbance measurements on the complete vehicle but also specific immunity tests for electrified and autonomous features. In the laboratories with which I have worked, we frequently refer to the national standards that cover vehicle-level EMC requirements, including GB 34660-2017, GB/T 18387-2017, and GB/T 37130-2018. These documents are often used together with international regulations, especially UNECE R10.06 for the uniform provisions concerning the approval of vehicles with regard to electromagnetic compatibility. There are also component-level standards, technology-specific standards, and OEM internal specifications that need to be understood in the context of an electric car project.
A summary that I have found useful when explaining the standard system to colleagues is shown in the following table. This table presents the main categories of EMC standards that affect the intelligent electric car and the general subject matter covered by each group.
| Category | Typical Standards | Purpose |
|---|---|---|
| Vehicle-level broadband and narrowband emission | GB/T 18387, CISPR 12, GB 34660 | Limits for electromagnetic radiation from the complete electric car in the 10 m test site |
| Vehicle-level immunity to electromagnetic fields | GB 34660, ISO 11451-2, ECE R10 | Verification of the vehicle immunity against radiated electromagnetic disturbance |
| Electrical transient conduction along supply lines | GB/T 21437.2, ISO 7637-2 | Test of electronic components against transient pulses on the power supply line |
| Coupling of electrical transients on signal lines | GB/T 21437.3, ISO 7637-3 | Test of electronic components against capacitive and inductive coupling on non-power lines |
| Electrostatic discharge | GB/T 19951, ISO 10605 | Evaluation of electronic systems under direct and indirect static discharge |
| Component-level conducted and radiated emission | GB/T 18655, CISPR 25 | Limits for voltages and electromagnetic fields generated by components |
| Component-level radiated immunity | GB/T 17619, ISO 11452-2 | Verification of components against electromagnetic fields in free-space and absorber-lined chamber methods |
| Charging communication protocol and EMC interaction | GB/T 18487.1, ECE R10.06 | Conducted disturbance on the AC supply line and radiation during charging |
One important thing I have learned is that compliance with the type-approval standard does not always guarantee the trouble-free operation of a complex electric car in the real environment. Standards represent a minimum set of requirements that has been agreed upon by regulators and industry committees. During development, however, the engineering team must usually implement additional verification steps that are much more demanding than the legal minimum. The most important value of a vehicle-level EMC test is not simply to check a box for certification; it is rather to reveal design weaknesses, coupling paths, software-state errors, and wiring problems that could affect customer satisfaction or safety. For this reason, I prefer to treat the standards as a basis for a more detailed test protocol that is specifically tailored to the technology under evaluation.
EMC Testing of ETC on an Electric Car
The first case that I want to present concerns the ETC system. An ETC installation usually consists of the on-board unit (OBU), the roadside unit (RSU), a host computer for transaction management, and additional devices such as cameras and automatic barriers. For the intelligent electric car, the OBU is usually attached to the windscreen, while the RSU is mounted on a toll gantry or similar roadside structure. The following table lists the main EMC test items for the OBU according to the national standard GB/T 38444-2019. This standard came into force on 1 July 2020 and defines test methods and test levels for the component under laboratory conditions.
| No. | Test Item | Reference Standard | Key Test Requirement |
|---|---|---|---|
| 1 | Electrostatic discharge | GB/T 28046.1-2011, GB/T 19951-2019 | Contact discharge, air discharge, and indirect discharge |
| 2 | Transient immunity along the supply line | GB/T 28046.1-2011, GB 34660-2017, GB/T 21437.2-2008 | Pulses 1, 2a/2b, 3a/3b, 4, and 5a/5b at level III |
| 3 | Electrical transient immunity through capacitive and inductive coupling | GB/T 28046.1-2011, GB/T 21437.3-2012 | CCC and DCC test at level III |
| 4 | Radiated electromagnetic immunity | GB/T 28046.1-2011, GB 34660-2017 | ALSE method and bulk current injection method |
| 5 | Conducted emission | GB/T 18655-2018 | Level III |
| 6 | Radiated emission | GB/T 18655-2018 | Level III |
The standard defines test methods at the component level, but it does not provide a detailed procedure for testing the complete vehicle with the OBU operating in its real transaction state on an electric car. During the front-loading phase of our development program, we therefore decided to perform a series of vehicle-level exploratory tests. We wanted to answer three questions. First, can we activate the OBU reliably inside a semi-anechoic chamber that is normally used for whole-vehicle emission and immunity tests? Second, what parameters should be monitored to determine whether the OBU has failed during the test? Third, how should the RSU antenna be placed without disturbing the vehicle test, particularly when an immunity antenna is radiating high field levels?
Work-Mode Definition and Performance Evaluation Criteria for ETC OBU
In our testing, the ETC OBU was found to have two typical operating modes. The first is the transaction mode, in which the OBU continuously communicates with the RSU for electronic toll collection. The second is the Bluetooth mode, in which the OBU is connected to a user’s mobile phone for account binding or configuration purposes. The transaction mode is the most relevant for toll collection on highways. The Bluetooth mode is important for customer setup. For the completion of an EMC test of an electric car, we must place the OBU in a deterministic state that remains stable and repeatable throughout the measurement. The following table shows the mode settings and the functional status criteria that I defined for the vehicle-level EMC verification.
| No. | Mode | Setup State | Judgement Criteria |
|---|---|---|---|
| 1 | Transaction mode | The OBU is continuously in ETC transaction status and never enters sleep mode; the transaction interval is set to 1 s. | Pass: the buzzer produces a short beep, the OBU is switched on, and the LED on the OBU shows green. Failure: the buzzer beeps three times or the LED changes to red. |
| 2 | Bluetooth mode | The OBU is switched on, the 2.4 GHz Bluetooth module is active and never goes to sleep, and the 5.8 GHz module continuously transmits a modulated signal with PN9 encoding. | Pass: the LED remains green throughout the test. Failure: the LED becomes extinguished or changes to red during the test. |
In order to maintain continuous operation of the ETC system, I used a host computer to configure the RF transmitter parameters of the RSU. The RSU output power and timing are crucial. If the output power is too low, the OBU may not be awakened reliably. If the transaction interval is too long, the test duration becomes excessive. In our experiments, we set the RSU transmitter to output a power level of 31 dBm at the antenna port, with a transaction interval of 1000 ms and an antenna gain of 10 dBi. The parameters are summarized in the following table.
| RSU Output Power | Transaction Interval | Antenna Gain |
|---|---|---|
| 31 dBm | 1000 ms | 10 dBi |
Radiated Emission Test Setup for the Vehicle with ETC OBU
For the whole-vehicle radiated emission test, I placed the electric car in the semi-anechoic chamber and positioned the RSU antenna at a distance of two meters in front of the windscreen. The RSU antenna height was set to about 2.5 m above the ground plane. The host computer was located outside the chamber, connected to the RSU controller by a control cable. The RF signal from the RSU controller was fed through a coaxial cable to the RSU antenna inside the chamber, thus producing a reliable wake-up signal for the OBU. The measurement antenna of the chamber received the electromagnetic radiation emitted by the vehicle, and the signal was delivered through an adapter to the receiver and then to the test computer.
During the first measurement campaign, the OBU was configured in the two states that I have described. The emission spectrum showed important disturbances in the 5.79 GHz and 5.83 GHz bands. At these frequencies, the signals produced by the OBU itself were far above the emission limits defined by the vehicle-level standard for commercial electric cars. The spectral curves indicated that the OBU generated radiation at its characteristic operating frequencies. According to the Chinese national regulation for ETC, the ETC system uses the 5.8 GHz DSRC frequency band. More precisely, the uplink frequency from the OBU to the RSU is 5.79 to 5.80 GHz, while the downlink frequency from the RSU to the OBU is 5.83 to 5.84 GHz. The receiving bandwidth of the OBU has a 6 dB bandwidth that reaches a maximum value of about 5.825 GHz to 5.845 GHz.
The measured over-limit peaks in the vehicle radiated emission test were exactly the characteristic operating frequencies of the RSU and OBU communication link. From a physical point of view, these frequencies are the intentional radiation of the ETC transceiver and therefore cannot be considered the same as spurious emissions originating from an electrical drive system in an electric car. After evaluating the regulations related to short-range radio equipment, we decided that these emission peaks could be exempted from the vehicle emission judgement, provided that the rest of the spectrum satisfied the applicable limits. This is an important lesson for the development of an electric car: the test engineer must carefully analyze the measured spectrum and separate the intentional radiation of communication modules from the unintentional disturbances that can impair the quality of radio reception in the car.
Radiated Immunity Test Setup for the Vehicle with ETC OBU
The situation becomes more complicated when we perform a whole-vehicle radiated immunity test on an electric car with the ETC system in operation. The immunity test antenna generates strong electromagnetic fields in the frequency range from 1 GHz to 6 GHz. If the RSU antenna is placed inside the main beam of the immunity antenna, the incident electromagnetic field will be coupled into the RSU antenna. This could disturb the RSU controller and interrupt the communication with the OBU. In order to maintain normal communication, I placed the RSU antenna behind the immunity antenna. In principle, this arrangement reduces the direct coupling of the high field into the RSU electronics while still providing a sufficient radio-frequency signal to wake up the OBU. During our verification, we found that when the RSU antenna was placed in a position where the vehicle was moved by 30 to 90 degrees relative to the normal axis, the transaction communication remained stable and reliable.
In a vehicle-level immunity test, we selected only the transaction mode for the OBU because this mode best represents the real operating environment on a highway. The monitored parameters and judgement criteria were consistent with the information shown in the preceding table. When I carried out the test, the OBU survived the radiated field without any functional failure, and the transaction behavior remained stable.
However, a different result was obtained when the same OBU was tested at the component level according to the OEM internal specification TL 81000-2018. The internal standard requires swept radiated immunity tests over a frequency range that extends from 1 GHz to 6 GHz, using both continuous-wave (CW) and pulse-modulated (PM) signal formats and both horizontal and vertical polarizations. The exploratory measurements on one OBU sample revealed abnormal transaction behavior in several frequency ranges. These anomalies are summarized in the following table, which lists the observed failure mode as repeated failure of the OBU to wake up, or repeated transaction failure during the exposure.
| Frequency Range / GHz | Modulation | Polarization | Observed Failure |
|---|---|---|---|
| 1 to 3.4 | CW | Horizontal | 12 wake-up failures in the range 2.06 to 2.13 GHz |
| 1 to 3.4 | CW | Vertical | 17 wake-up failures in the range 2.05 to 2.17 GHz |
| 3.4 to 6 | CW | Vertical | One wake-up failure at 3.72 GHz |
| 3.4 to 6 | CW | Horizontal | 3 transaction failures and 47 wake-up failures in the range 5.6 to 6 GHz |
| 3.4 to 6 | PM | Vertical | 1 transaction failure and 109 wake-up failures in the range 3.4 to 6 GHz |
| 3.4 to 6 | PM | Horizontal | 8 transaction failures and 120 wake-up failures in the range 3.4 to 6 GHz |
A careful analysis of the results showed that the frequency range from 2.05 GHz to about 2.17 GHz corresponds to the downlink frequency bands used by CDMA2000 and WCDMA mobile communication systems in China. In the real road environment, the base stations for these systems are usually located far away from the ETC toll gantry. The field strength generated at the windscreen of an electric car while passing through the toll station is therefore much lower than the test level applied in the component-level immunity test. The probability that these signals will disturb the OBU during the actual transaction is extremely low. Therefore I recommended that, for the same OBU, the frequency range from 2 to 3 GHz could be exempted from the component-level and vehicle-level radiated immunity judgement when the system has been shown to comply with the applicable national ETC standard.
I also evaluated the hand-operated mobile-phone disturbance test described in some OEM standards. The internal vehicle-level standard defines several positions where a mobile phone must be placed during the test, such as on the seat, in the door storage well, near the central display, and close to the cable ports of the device under test. For the ETC OBU mounted near the interior rear-view mirror, the real distance between the OBU and the mobile telephone placed at the dashboard is normally larger than 40 cm. The radiation from the mobile telephone at such a distance has almost no effect on the OBU performance. In our exploratory test, the actual result was found to be acceptable. It is therefore not necessary to carry out the hand-phone test at the vehicle level for the OBU installation.
Reflections from the ETC Case
The ETC case taught me that testing the complete electric car is not just a matter of applying the regulatory standard blindly. We must carefully distinguish between intentional radio emissions and interference, choose a transaction mode that is representative of normal use, place the roadside antenna so that it does not become a coupling path for the immunity field, and establish a set of functional monitoring criteria that can be observed remotely from outside the chamber. The same logic can be extended to other radio-frequency systems in an electric car, including satellite navigation, broadcast receivers, tire-pressure monitors, remote keyless entry, and vehicle-to-everything communication modules.
EMC Testing of the Charging Function for Electric Cars
The second case in my research concerns the electromagnetic compatibility of the charging process for an intelligent electric car. Charging is one of the most characteristic operations that distinguishes an electric car from a conventional vehicle. During the charging state, the power electronics inside the vehicle are active, high switching currents are flowing between the battery and the on-board charger, and the vehicle is physically connected to the external electricity grid. The electromagnetic noise produced by the traction battery charging process can be coupled through the charging cable and may disturb other electronic systems in the immediate vicinity. Conversely, external disturbances on the power grid can enter the vehicle through the charging port and affect the on-board electronics.
The international regulation ECE R10.06 contains specific requirements for the electromagnetic compatibility of vehicles in the charging condition. The following table summarizes the tests that are relevant to vehicle charging and the measurement methods that I normally apply.
| No. | Test Item | Reference Standard | Remarks |
|---|---|---|---|
| 1 | Conducted radio-frequency interference on the AC or DC supply line | CISPR 16-based methods, CISPR 12 adapted | Frequency range from 150 kHz to 30 MHz |
| 2 | Broadband electromagnetic field radiation from the vehicle | CISPR 12, CISPR 22 | Usually measured at a test distance of 10 m |
| 3 | Narrowband electromagnetic field radiation from the vehicle | CISPR 12, CISPR 25 | Usually measured at a test distance of 10 m |
When I measured the radiated emission of the electric car during charging, the test arrangement seemed straightforward at first. The car is parked in a semi-anechoic chamber, the AC charging station is placed outside the chamber or in a specially protected area, and the charging cable is passed through a waveguide or a shielded access panel into the chamber. The vehicle is then put into the charging state. However, a serious difficulty appears when we try to isolate the noise generated by the on-board charger from the background noise produced by the external charging station. The AC charging pile contains rectifiers, switch-mode power converters, filters, and control electronics. These components generate electromagnetic noise that is injected into the vehicle through the charging cable. As a result, the measured emission is a combination of vehicle noise and charger noise. It is difficult to determine the contribution of the on-board charger alone.
Another problem appears after an immunity test has been completed. It is necessary to verify the charging function of the vehicle quickly and objectively. A conventional visual inspection is not enough because the charging process depends on a complex handshake between the electric car and the charging station. A fault may appear only when the control pilot signal is not recognized correctly. For these reasons, I decided to design a special test device that could solve both tasks.
Dedicated Test Device for Charging EMC Measurement
The proposed test device contains two functional sections. The first section is a vehicle-load simulator. This simulator is capable of reproducing the vehicle recognition signal required for alternating-current charging according to the standard GB/T 18487.1-2015. It also supplies the proper termination resistance that the charging station must detect when it is connected to a vehicle. In this way, the simulator creates the same electrical conditions as an electric car at the charging outlet but does not introduce additional electromagnetic noise. When we connected a real charging pile to this simulator, the charging pile entered its normal state and supplied the correct voltage to the vehicle-side equipment. We could therefore use the arrangement to measure the background noise of the charging pile alone. After that, we connected the charging pile to the on-board charger of the electric car through the simulator, and the complete noisy environment became representative of the actual electric car charging state.
The second section of the device is a functional verification module for the on-board charger. The module is used after the vehicle has been exposed to fast electrical transients, surge pulses, or other disturbances that are coupled along the vehicle power supply path. The module communicates with the vehicle through the charging control pilot line and sends the handshake protocol defined by GB/T 18487.1-2015. If the electric car successfully performs the charging preparation procedure, the on-board charger is considered to be functioning normally. If the vehicle diagnostics report a fault, or if the charging preparation is not completed, the test is regarded as a failure. The practical test result of the handshake signal was very stable, which demonstrated that the module could correctly activate the vehicle charging state.
The introduction of this dedicated device brought two great benefits. First, it allowed us to obtain a clean background spectrum from the charging pile and then measure the incremental noise caused by the on-board charger of the electric car. This separation of noise sources is extremely helpful when the vehicle emission exceeds the limit and we need to determine which side of the charging interface is responsible. Second, the functional verification module provided a fast and repeatable way to confirm that the vehicle’s charging function had survived an immunity test. The former procedure that relied on a human operator and a real charging station was both slow and subjective. With the new device, the function check can be completed within a few seconds and the result is unambiguous.
Observations from the Charging Test Campaign
Through a series of charging EMC tests, I observed that the switching frequency of the on-board charger and its harmonics could sometimes appear in the conducted emission spectrum with a magnitude that was close to the regulatory limit. The main countermeasure that proved effective was an improvement of the filtering network inside the charger. However, the test device also highlighted a phenomenon in which the cable layout between the AC charging inlet and the charger had a strong influence on the emission level. Because the charging harness passes through the vehicle underbody, its length and the shielding status can create unanticipated resonances. I therefore recommend that the EMC performance of an electric car charging system be considered at the same time as the high-voltage harness routing is designed, not as an afterthought in the final validation phase.
EMC Testing of the On-Board Communication Unit for Intelligent Electric Cars
The third case in my article is the on-board communication unit, often abbreviated as OCU. This unit is the gateway through which the intelligent electric car sends its position, status, emergency calls, and remote service information to the cloud. The OCU must be in continuous contact with mobile communication base stations, and it may also receive signals from the global navigation satellite system. Because these radio functions are essential to the connected vehicle, the electromagnetic compatibility test of the OCU deserves special attention.
The standard GB/T 32960.2-2016 “Technical specification for remote service and management systems for electric vehicles, Part 2: Vehicle terminals” defines the EMC requirements for the vehicle terminal. The relevant test items are listed in the following table.
| No. | Test Item | Reference Standard | Key Test Requirement |
|---|---|---|---|
| 1 | Electrostatic discharge | GB/T 28046.1-2011, GB/T 19951-2005 | Contact discharge +/-6 kV, air discharge +/-15 kV, performance class A |
| 2 | Transient immunity along the supply line | GB/T 21437.2-2008 | Pulses 1, 2a/2b, 3a/3b, 4, 5a/5b at level III |
| 3 | Coupled electrical transient immunity | GB/T 21437.3-2012, GB/T 28046.1-2011 | CCC test at level III |
| 4 | Radiated electromagnetic immunity | GB/T 17619-1998, GB/T 28046.1-2011 | Free-field method and bulk current injection method |
| 5 | Conducted emission | GB/T 18655-2010 | Level III |
| 6 | Radiated emission | GB/T 18655-2010 | Level III |
At the component level, the OCU can be tested with a dedicated RF link because its antennas can be connected to a communication tester through coaxial cables. At the complete vehicle level, however, the test becomes much more difficult because the vehicle has the integrated antennas, wiring harness, and body structure of the electric car. The vehicle-level radiated emission test would normally require the OCU to be active, receiving real signals from a mobile network simulator, and transmitting at its maximum output power. This often requires significant laboratory investment. In one project, the development team proposed to upgrade the semi-anechoic chamber by installing filtering interfaces and dedicated antennas inside the chamber so that an external base-station simulator could communicate with the OCU. Although this solution provides excellent accuracy, it requires a large amount of capital expenditure and involves modification of the shielded chamber, which may create a risk of compromising the shielding effectiveness.
In order to avoid this large investment, I developed an alternative low-cost method for vehicle-level radiated emission testing of an OCU on an electric car. The method is based on the idea of conducting the telecommunication signal into the chamber through an RF cable rather than radiating it through an internal antenna. The OCU is connected by a short RF cable to the vehicle antenna port. The cable passes through a waveguide port in the chamber wall and reaches an external receiver antenna or a telecommunication simulator. The arrangement activates the OCU by simulating the communication signals from a base station, while the electromagnetic emissions of the OCU itself are measured by the normal receiving antennas of the chamber.
There are two technical problems that need to be solved in this approach. The first one is the shielding effectiveness of the waveguide port. If the RF cable or the waveguide is not correctly designed, external electromagnetic noise from the environment can penetrate into the semi-anechoic chamber and affect the measurement. The second problem is that the vehicle antenna of the OCU is removed from its original position, so we need to provide an external receiving antenna with a performance that is similar to that of the real vehicle antenna.
For the waveguide port design, I used the following classical equation for the cutoff frequency of a circular waveguide:
$$f_c = \frac{1.76 \times 10^{10}}{d}$$
where \(f_c\) is the cutoff frequency expressed in GHz, and \(d\) is the diameter of the waveguide opening expressed in centimeters. The value \(s\) represents the shielding attenuation factor in dB, and \(L\) is the length of the waveguide in centimeters. To ensure that the OCU communication band is not disturbed by external signals, the cutoff frequency should normally be more than five times higher than the frequency of the signals that need to be shielded. The OCU used in the project works at frequencies below 3 GHz. I therefore set the cutoff frequency at 15 GHz. Substituting this value into the equation gives an aperture diameter of approximately 1.16 cm. In order to maintain a shielding performance of more than 100 dB, the waveguide length had to be about 40 cm. The dimensions are summarized by the expression
$$L \approx \frac{s}{0.0205 f_c \sqrt{1 – (f/f_c)^2}}$$
where \(s\) is the desired shielding attenuation in dB, \(f_c\) is the cutoff frequency in GHz, and \(f\) is the operating frequency. In practical application, I filled the area around the extended RF cable in the waveguide chamber with a high-density copper wire mesh. The clearance between the mesh and the RF cable, as well as the length of the mesh inside the aperture, was selected according to the above calculation. The resulting chamber port has been used in many subsequent tests without a noticeable increase in the noise floor.
For the receiving antenna issue, I designed a dedicated test plate with an area of 200 mm × 200 mm. The purpose of this plate is to simulate the mounting environment of the real vehicle antenna. I then measured the reflection coefficient of this test antenna and compared it with the response of the real antenna mounted on the roof of the electric car. The simulation results showed that the \(S_{11}\) parameters of the two arrangements were very similar over the frequency range of interest. The design was therefore considered suitable for the vehicle-level test.
During the whole-vehicle radiated emission test, the operating mode of the OCU was set to the emergency-call mode, which caused the unit to transmit periodically at its maximum power. The measured spectrum is shown in the form of the final experimental result. The main over-limit peaks that appeared in the spectrum were found to be the characteristic operating frequencies of the OCU itself. These peaks have a narrow bandwidth and occur at frequencies where the OCU is intentionally radiating to communicate with the mobile base station. Therefore, they can be exempted from the vehicle emission limit, just as we did for the ETC system. It is important to note that no extra interference was introduced by the new measurement setup. The background noise remained low, which confirmed that the shielded waveguide port and dedicated antenna plate had been correctly constructed.
From the perspective of a test engineer, this low-cost vehicle-level method has several advantages. It avoids expensive chamber modification. It maintains the high shielding performance of the semi-anechoic chamber. It allows the OCU to be activated from an external communication tester without installing a separate antenna mast inside the chamber. Furthermore, it supports not only radiated emission measurement but also several types of immunity and functional tests, because the RF link to the OCU is available through a stable conductive interface.
Practical Test Setup and Uncertainty Considerations
When I now plan the EMC experiments on an intelligent electric car, I always consider several practical aspects that are not always mentioned in standards but have a great influence on measurement accuracy. The first is the need to place the test antennas at the correct height and distance. This applies particularly to the 5.8 GHz ETC band, where small errors in the antenna position can change the coupling by several decibels. The second is the need for common-mode rejection on the charging cable. A long charging cable can act as an antenna and increase the apparent radiated emission of the electric car. The third is the need to use optical links or battery-powered data acquisition systems when we need to monitor the vehicle status without running long cables that could disturb the field.
Another important consideration is the grounding of the battery charger or charging pile used in the test. In several cases, I observed 50 Hz hum and broadband noise in the emission spectrum that was not caused by the vehicle but by the poor grounding of the charging station. The dedicated vehicle load simulator helped me to avoid this problem because it created the necessary electrical state without adding new emissions. A similar approach can be used for other components of the electric car, such as the DC-DC converter, the electric air-conditioning compressor, and the auxiliary heater. In every case, the background noise must be measured and documented before the final vehicle measurement is made.
I also use the following table in my everyday engineering practice to summarize the evaluation of measurement feasibility for intelligent electric car subsystems. This table is intended to help project teams to understand which test approach is suitable for a given function and what kind of test equipment is needed.
| Subsystem | Component Test Level | Vehicle Test Level | Special Equipment | Main Judgement Parameter |
|---|---|---|---|---|
| ETC OBU | Required by GB/T 38444 | Exploratory according to OEM method | RSU with external antenna and controller | Transaction success, LED state |
| OCU / vehicle telematics | Required by GB/T 32960.2 | Vehicle-level radiated emission method | Waveguide port, external RF antenna, communication tester | Transmission state, communication link, LED state |
| On-board charger and charging pile | Not fully standardized for separate charger | Required by ECE R10.06 and GB standards | Vehicle load simulator, functional handshake module | Charging preparation status, handshake signal, emission spectrum |
The table has served as a practical guide for many of my colleagues. It emphasizes that the test items, test methods, and evaluation criteria vary from one technology to another. It is impossible to use one single uniform procedure for all functions that appear in the modern intelligent electric car.
Challenge Cases and Engineering Solutions
In the course of my daily EMC work, I have encountered several situations where the component-level test result did not agree with the vehicle-level test result. One prominent example was the ETC OBU. The component-level radiated immunity test according to a strict OEM specification produced multiple failures in the 2 GHz and 5.8 GHz bands, while the vehicle-level test with a realistic field strength showed no degradation. This can be explained by the antenna factor, the position of the OBU on the windscreen, the shielding effect of the vehicle body, and the different field distribution between the component test chamber and the vehicle test site. A decision based solely on the component-level result would have led to an unnecessary redesign of the OBU. A decision based only on a vehicle-level test at a low field strength would have hidden a real weakness in the 2 GHz band. The final judgement therefore required a risk analysis that considered the actual electromagnetic environment around an ETC toll station.
A second example was the conducted emission measurement during charging. The AC charging pile that we initially used produced a relatively high background noise. The measured spectrum of the electric car was very close to the limit, but after switching off the charger and measuring only the vehicle with the simulator, we discovered that the main harmonics were actually produced by the charger and not by the vehicle. This demonstrates the value of redundant diagnostic methods: the design of the test is as important as the test equipment itself.
A third example involved a telematics unit that showed excessive radiated emission at the vehicle level. The narrowband peaks were originally assumed to be the intentional emission of the mobile communication module. After a closer investigation, I found that one of the peaks was not related to the communication standard at all. It was an oscillation in the power supply circuit of the infotainment system, generated when the OCU was transmitting at high power. The coupling path was the shared power supply harness. This problem could not be identified from a component-level OCU test alone. The full vehicle test with the new RF activation method provided the necessary visibility into system-level interactions.
Test Results, Data Analysis and Judgement Methodology
The evaluation of EMC test results for an electric car must always take into account the appropriate frequency bands and the physical operation of the wireless systems. In many regulatory tests, the limits are defined for the protection of broadcast and radio services. Therefore, the transmitters of an ETC system or a mobile communication unit should not be treated as unintended disturbances. In the standard practice that I use in my laboratory, the judgement of radiated emission is divided into two parts. In the first part, I measure the electromagnetic radiation of the electric car with the transceiver in its normal state. In the second part, I compare the measured spectrum with the frequency plan of the radio system. If the over-limit peak coincides exactly with the intentional transmission band and has a bandwidth that is narrower than the channel bandwidth of the system, I record it as the expected intentional emission. If a peak appears outside the intentional transmission band, I investigate the cause in more detail.
The functional status monitoring is also criticial for the interpretation of immunity tests. I use the following basic principles. First, the monitored function must be observable from outside the test chamber. This is usually achieved through optical cameras, optical fiber links, or wireless video transmission. Second, the function must be monitored with a resolution that is sufficient to detect short interruptions. A miss of only a few milliseconds may indicate a susceptibility issue that could lead to a transaction error in an ETC system. Therefore, I monitor several status events, not merely the final voltage or communication state. Third, the test event log must be synchronized with the electromagnetic disturbance. In my own setup, I record a timestamp for each transaction and compare it with the time axis of the EMC measurement system. This allows me to say exactly which frequency caused an interruption.
Future Development of EMC Testing Methods
The field of intelligent electric cars is expanding rapidly. The number of wireless services in a vehicle is growing, and the number of power electronics modules is also growing. The interaction between the high-voltage battery system and the communication network has become an important topic. In the future, I expect that vehicle-level EMC testing will be complemented by more advanced methods such as reverberation chamber testing, mode-stirred measurements, and over-the-air testing of complete vehicles. These methods are able to emulate realistic multi-path propagation and support the validation of antenna systems installed on an electric car.
The challenges also come from the software complexity of the vehicle. Modern electronic control units have many operating modes, software updates, and fail-safe strategies that can change their electromagnetic behavior. A test engineer cannot simply switch on a device and expect it to perform in exactly the same manner as at the beginning of the development. Therefore, I recommend that the EMC team works together with the software team to define a set of stable operating conditions for each EMC test. The definition should include the communication bus data, the radio transmission status, the charging state, and the state of every relevant actuator.
An additional challenge is the electromagnetic field distribution around the wireless charging systems used for electric cars. Although the present article has focused on conducted and radiated EMC tests with external charging equipment, I would like to emphasize that the same systematic approach should be followed when evaluating wireless inductive charging devices. The high-frequency magnetic field generated by the ground-side coil may interfere with the vehicle’s own keyless entry system, tire-pressure monitoring, and mobile communication links. In this case, dedicated modeling tools and a well-defined measurement procedure are necessary to quantify the interference and design countermeasures.
My experience also suggests that cost-effectiveness will continue to be a major concern. A full vehicle EMC chamber is an expensive investment. Laboratory time is always limited, especially during winter tests and pre-production validation. In this context, a carefully prepared test setup that is flexible enough for different subsystems can save many hours. The dedicated vehicle load simulator for the charging test and the waveguide-port RF activation system for the OCU are examples of this philosophy. They prove that innovation in test methods is often as valuable as innovation in the product itself.
Conclusion
In this article, based on my practical work in the field of electromagnetic compatibility, I have described several research activities and test practices for intelligent electric cars. The main emphasis has been placed on the complete vehicle test approach, the detailed preparation of test arrangements, and the interpretation of measured data. The first case study concerned ETC equipment and showed how the RSU antenna, the communication transaction mode, and the vehicle-level judgement can be arranged to ensure repeatable tests inside a semi-anechoic chamber. The ETC over-limit emission at 5.79 GHz and 5.83 GHz was analyzed as intentional radiation, and an exemption logic was defined. The immunity test of the ETC OBU, on the other hand, showed some sensitivity in the 2 GHz mobile communication band. I proposed that this band can be treated as a negligible risk because the distance between the toll gantry and mobile base stations is large enough to suppress possible interference.
The second case study considered the EMC of the charging process for an electric car. I introduced a dedicated test device that contains two main modules. The vehicle load simulator allows the charging pile to work normally without adding electromagnetic noise, which is essential for the accurate measurement of the vehicle’s charging emission. The charger function verification module sends the correct handshake protocol through the control pilot line and provides a fast pass-or-fail method after transient tests. This device has solved a significant practical problem in my laboratory and has improved the quality of our charging EMC evaluation.
The third case study concerned the OCU test at the vehicle level. I described a low-cost method based on a shielded waveguide port and an external RF communication antenna. The design was supported by a cutoff-frequency calculation that ensures high shielding effectiveness for the chamber. The dedicated antenna plate produced an \(S_{11}\) characteristic that was comparable with the real vehicle antenna. The method successfully activated the OCU without introducing additional electromagnetic disturbance and allowed us to perform the vehicle-level radiated emission test on an intelligent electric car.
All these cases lead to one important conclusion. Electromagnetic compatibility testing of an intelligent electric car is not a static task that can be fulfilled by copying a standard procedure from an old vehicle program. It requires a deep understanding of the technology under test, the physical mechanisms of coupling, the frequency spectrum occupancy of intentional transmitters, and the actual electromagnetic environment in which the vehicle operates. The methods that I have presented in this article are intended as an engineering reference for those who are facing similar test problems in their own development work.
Looking ahead, I believe that the EMC discipline must continue to evolve in parallel with the development of electric powertrains, intelligent driving systems, and cloud-connected mobility. The test techniques that work today may become obsolete tomorrow. However, the underlying logic remains the same: we must define a repeatable state, apply a well-characterized disturbance, monitor a functionally relevant parameter, and separate intentional emissions from unwanted noise. This logic will remain the core of automotive EMC testing regardless of whether the vehicle is powered by an internal combustion engine, a battery, a fuel cell, or any other technology that will appear in the future.