How to test EMI capacitors for xEV systems
IEC?60384-14 is the relevant test standard for EMI?capacitors. If used in automotive applications, the AEC-Q200 must also be observed. When it comes to xEV systems, for some of the required tests, IEC 60384-14 is more appropriate, for others, however, AEC-Q200. This article provides guidance on which tests can be required for xEV applications and which standard should be used for each test.
by David Olalla , Vice President of Product Development, Film Capacitors Industry & Automotive, and
Felipe Oliveira , Product Manager Industry, Film Capacitors Industry & Automotive, TDK Electronics
Capacitors for EMI suppression filter undesired conducted transients to or from the power source and to or from surrounding systems. IEC?60384-14 (Fixed capacitors for electromagnetic interference suppression and connection to the mains) has been the main reference for testing and certification of this sub-category of capacitors since the last century. It is commonly required in a variety of end applications: LED lighting, consumer electronics, solar inverters, power supply units, etc., and nowadays also increasingly in certain xEV systems that are connected to a power grid, such as on-board chargers, DC-DC converters, and filters.
IEC?60384-14 is mainly targeted to applications connected to the AC?grid with 50/60?Hz as the primary harmonic and perhaps some smaller superimposed higher frequency harmonics but can also be connected to other power sources such as batteries or solar cells if certain safety requirements are to be met. Sometimes, EMI capacitors are found in output filters downstream of the inverter, which have a higher content of high-frequency harmonics. This is an example of an aspect not explicitly addressed by IEC?60384-14.
AEC-Q200, on the other side, is a reference standard and stress test qualification for passive components in automotive applications. Although it contains methods and requirements for testing film capacitors, it lacks specific information about the system, the position of the circuit, or the permissible drifts. The user of the components should provide this information in a separate specification.
Some users in the xEV segment may not need full IEC?60384-14 certification because it is simply not required for their specific system or because it is not connected to the power grid. However, they may require a combination of AEC-Q200 compliance and some specific IEC?60384-14 safety tests (e.g., high voltage test, impulse voltage test) to ensure that the capacitor can properly handle the expected load of the overall system when tested to higher standards.
In summary, three approaches for EMI capacitor qualification can be applied to xEV systems:
In addition, it is not only necessary to check whether the tests have been correctly selected following the safety tests at the board level and the test requirements for the end application. Furthermore, whether the specified parameter drift is acceptable must be checked for each test. This ensures the proper operation of the entire system during its lifetime.
Limitations of the standards
Table 1 provides a relevant selection of the test criteria for film capacitors from both standards and compares them with each other. One first difference is that the AEC-Q200 refers to the specification of the supplier or user, while IEC 60384-14 refers to fixed limit values. This is not a contradiction, as the specification of the (capacitor) supplier usually refers to the IEC criteria (or even stricter ones).
In addition, capacitor users must also consider how these specific drifts and limits affect their systems, as the following examples show:
The limit values for insulation resistance and tan?δ specified in IEC?60384-14 have a general scope and apply to different capacitor technologies, materials (e.g., ceramic, polypropylene, polyester, paper, etc.), and capacitance ranges. However, properties such as dielectric losses, leakage current, or temperature coefficient differ in practice depending on the technology. Therefore, it is necessary to check carefully to what extent the capacitor's performance applies to the respective tests.
Firstly, suppose a particular test is representative or a test with accelerated stressors is feasible. In that case, the actual parameter drift, in this case, must be thoroughly re-examined to consider the consequences of this drift in the real application. For example, parts of the qualification tests mentioned above (outliers or maximum values) can be used and tested in the real system to validate the robustness.
Secondly, a test itself might be so demanding that it induces different failure or aging mechanisms than those that occur in practice. In that case, the effects of these drifts should not be considered so extensively in the real application. A good approach is to compare the drifts in the present test with those in previous ones with similar components. Does the capacitor clearly behave worse? If so, the issues and possible effects on the application should be examined very carefully.
Temperature cycling
Five temperature cycles, as required by IEC?60384-14, are inadequate for automotive applications. In contrast, one thousand cycles in the AEC-Q200 shock test from -55?°C to +85?°C or a maximum temperature category with a one-minute transient is a well-established accelerated test for capacitors in such applications. Considering the real-world applications, this seems to be very challenging as the mass of the overall system dampens the temperature swings in these small components. Nevertheless, this approach is well suited to evaluate the behavior of different materials during rapid temperature changes based on historical test data. This is useful, for example, when testing novel materials (housing, terminals, or dielectric) or when testing the contact between dielectric (polymer film) and metals (electrodes and terminals).
The example in Figure?1 shows the test results of an established product for automotive applications (TDK B3267*P*, 630?V). The dissipation factor tan?δ increases only slightly and is well below the mandatory limit values.
The example in Figure?2 compares three TDK film capacitors for 1000 cycles from -55?°C to +110?°C (AEC-Q200), showing a very stable performance for three of them (more stable than those specifically developed for automotive applications: B3203*AB* and the latest B3202*H*).
Vibration test
The vibration test is carried out at the component level, whereby the PCB and the mounting do not necessarily have to be the same as the final system. Consequently, it does not represent the conditions to which the capacitor may be exposed in the field. Nevertheless, obtaining a basic assurance of its performance is necessary, e.g., whether the connections are robust enough. In this context, the performance is always compared with empirical values from products already proven themselves in automotive applications.
Humidity test
With a duration of 1000?h, this AEC-Q200 test is located exactly in the middle between humidity grade?I (504?h) and grade?II (1344?h) of IEC?60384-14. The severity?S of a humidity test can be defined as the product of the two factors ρ and FD. Here, ρ describes how much water vapor is contained in the air and is a function of the relative humidity and temperature, and FD is the thermal acceleration factor for diffusion.
In a real application (in an open environment), the relative humidity decreases with increasing air temperature. Therefore, scenarios such as +85?°C/85%?RH (relative humidity) can only occur in a sealed enclosure with trapped humidity (e.g., in a test chamber) or for such a short time that diffusion from the cabinet to the environment does not occur.
Therefore, +85?°C/85%?RH (Figure?3) is suitable for assessing how robust an enclosure is against moisture. With an EMI capacitor, however, this leads to much more serious failures than in reality, especially if voltage is also applied. This is illustrated in Figure?4 using an X2?series from TDK.
Instead, +40 °C/93% RH would be a very common scenario in the field. In such a case, the parameter drift and its effects must be considered. When voltage is applied, the probability of partial discharges increases exponentially, more so with AC than with DC voltage.
So, suppose this test aims to simulate real-world conditions and accurately verify parameter drift and its effects on the application. In that case, it is also important to use AC or DC voltages as expected in the application and to consider average voltages (not accelerated) so as not to provoke other failure mechanisms.
领英推荐
Endurance test
The endurance test with stress factors like voltage and temperature is used to simulate and fundamentally verify the intended stress in the application. Parameter drifts and how they affect the application should be carefully checked.
With a voltage factor of 1.25 for X?capacitors, IEC?60384-14 applies the same approach as the AEC-Q200. A factor of 1.7 for Y?capacitors (Fig.?5), according to IEC?60384-14, is intended as a safety test (and not as a lifetime simulation). This is because Y?components are critical and usually connected to protective earth and, therefore, to touchable parts. Failures can lead to serious incidents.
Regarding temperature, the AEC-Q200 defines a value of +85?°C (or higher, depending on the user and supplier) as a reference for film capacitors. In comparison, IEC?60384-14 uses the maximum temperature of the category (usually +105 °C, +110 °C, or even +125 °C). Again, IEC?60384-14 follows a safety-oriented approach (testing under worst-case conditions), while the AEC-Q200 focuses on the expected application conditions with a certain acceleration factor.
How film capacitors age depends on the temperature and voltage. Therefore, a voltage acceleration factor (above 1.3) that is too high combined with a very high temperature could lead to excessive parameter drifts or test failures that would not appear in real-life applications. Figure?6 shows an example of a test at +125?°C with an acceleration factor of 1.1.
Flammability test
IEC?60384-14 focuses on component performance concerning active and passive flammability, while AEC-Q200 focuses on enclosure material class as per UL?94.
The result is simple: pass or fail. For this test, parametric drift is not applicable.
Soldering test
The standard test for heat resistance during soldering (+260?°C for 10?s) is based on severity requirements and criteria that are generally not critical for film capacitors unless they are very small.
However, the user must conduct this test on a real circuit board. This is the only way to ensure that the wave soldering process does not cause overheating and does not damage the film capacitor (Figure?7). In such a case, the tan?? would drift excessively, or in the worst case, the side areas of the housing would deform near the solder connections.
Requirements in xEV systems
Three applications where EMI capacitors are usually used in xEV systems are the motor inverter, the DC-DC converter, and the on-board charger (OBC). EMI capacitors are primarily used to suppress unwanted harmonics and provide dielectric strength to pass safety tests according to higher-level standards.
The power supply for the motor inverter and the DC-DC converter is the high-voltage battery and, therefore, DC. Although IEC?60384-14 primarily targets AC?applications, this standard must be referred to due to the necessary safety requirements. This applies particularly if any parts are connected to power lines or ground. IEC?60384-14 Edition?4 contains an annex that deals with the nominal values for DC?voltage and the required qualification tests. The requirements may differ as some users require higher test voltages, AC or DC tests, shorter or longer test times, leakage current measurements, or even a certain number of repetitions without damaging the components or degrading their service life.
High-voltage testing with AC is much harder than testing with DC. One of the reasons is that the impedance of a DC source is usually higher. Another reason is that the capacitance values are either higher or more capacitors are connected in parallel. This is because, in the case of self-healing during extreme overvoltage events or flashovers in the clearances, the impedance of the overall system under test is lower, and therefore, larger metalized areas evaporate.
Figure?8 gives an example of good design practice for high-voltage testing. It shows the results of three different product series used in xEV?systems as EMI?filters connected to the ground when subjected to a test level far above the specification. IEC?60384-14 states that the safety test requires 1500?V?(AC) for 60?s.
This is a real-world example of a combined test that goes far beyond the specification and is not included in any standard. This allows the user to select the right product depending on high-voltage requirements and other variables such as size, cost, or operating conditions. First, a thermal test was conducted to simulate the thermal stress of soldering PCBs. Next, a high-voltage AC test was performed on the system's whole set of capacitance and capacitors (worst case). Finally, possible repetitions of the high voltage along the power supply chain were simulated by gradually increasing the test time. Combinations?B and C in Figure?8 performed better in this series of tests, while A was finally selected based on the best combination of size and performance.
All three product series meet the requirements of AEC-Q200 and pass the high-voltage test according to IEC?60384-14 for Y2?capacitors, including a sufficient safety margin.
As mentioned in the previous example and the following real-life example, it is important to consider the total capacitance and capacitors of the entire system when subjecting it to high-voltage testing (Figure 9). While the supplier must provide information on an individual capacitor, it is still necessary to perform tests at the system level.
Other stressors
As outlined in IEC?60384-14 and AEC-Q200 (see Table?1), the main stressors are voltage, temperature, and humidity. However, the user should also consider other factors to select the most suitable component for the expected requirements before performing a qualification test.
A good example is the voltage slew rate (dV/dt), which can cause a high peak current (Ipeak) in conjunction with the capacitance. If this current peak is too high and is linked to high voltages or high temperatures, it can damage the internal contacts of the capacitor. If the dV/dt is high, the ESL (equivalent series inductance) must, therefore, be evaluated, as this causes the voltage to overshoot and other components to be stressed. SPICE models valid in the frequency and time domain are useful tools for simulating all these effects in the overall circuit.
The ripple current (inverter switching frequencies and harmonics) causes thermal losses heating up the capacitor and its surroundings. Combined with high voltage and temperature, this can result in thermal runaway of the component. None of those international standards mentioned above consider such an evaluation of the thermal behavior.
TDK Offers Components and Tools
The relevant IEC standards and AEC-Q200 conformity are summarized in detail in the TDK data sheets. Tests and limit values are well described and users can access them anytime. In addition, TDK offers online simulation tools (CLARA, Capacitor Life And Rating Application; Fig. 10) and complex SPICE models valid for the time and frequency domain. TDK thus supports its customers in selecting the most suitable capacitor for each application, thus avoiding oversizing or undersizing. Everything can be found on the TDK website.
The reference standards combined with the most diverse requirements from the application are constantly challenging capacitor manufacturers, leading to ever-new series developments and market launches.