Frankonia EDTC-AX

Introducing the Frankonia EDTC-AX E-Drive Testing Chamber defined and designed for powertrain tests in axis setup.

Technical measures to fulfill practical and CISPR 25 compliant EMC tests on e-axles

EMC tests of e-drives and e-axles at component level according to CISPR 25 require technical measures and implementation strategies to achieve realistic measurement results. Due to the increasing demands on performance classes, a realistic installation and implementation for e-axles, as well as the resulting mechanical requirements for a test bench itself are challenges. In relation to the size of the test object, a significantly larger test bench must be considered. This results in an increase in metal structures in the defined measurement environment (anechoic chamber). It is also important to note that, depending on the test bench type, the cable harnesses are routed close to the metallic structures of the test bench, ideally to the length required by CISPR 25 on the table ground plane, which leads to further challenges.

The Long Wire method was used to investigate the influences on the interference emission range from 150 kHz to 1 GHz for various test set-up variants in practice. This study is intended to stimulate a discussion regarding the comparability and reproducibility of measurement results on different test benches for electric drives or electric axles and to encourage further investigations.

?

Normative basics

A set-up for e-axles is according to CISPR 25 normatively not defined. However, influencing parameters for the different test bench concepts are important. For e-axle tests at component level and the idea of replicating the set-up as realistically as possible to the actual installation as in the vehicle, system-level tests based on CISPR 12 (vehicle) may be realistic.

Possible set-up variants and test parameters

According to CISPR 25, there are various practical design variants that contributed to the investigation, as well as the so-called Long Wire method, which was used as a measurement basis to evaluate the set-up.

• Cable harness routed close to metallic structure
• Use of ferrites to reduce reflections
• Antenna routed close to metallic structure
• Feeding in 120 dBμV in the range 150 kHz to 1 GHz
• Measurement of the emission including antenna factor
• Determination of the difference compared to the reference level from CISPR 25
• Requirement: 90% of all points within ±6 dB band

?

Measurements and results

Case 1: Change in the metal structure above the table ground plane

It has been verified that the measurement results deteriorate when metal structures are placed on the test table. This means that the influence of metal above the ground plane affects the measurement results.

?

Case 2: Influencing the built-up metal structure by using isolation material and ferrites

If the metal structures are installed on isolation from the table ground plane, the result < 130 MHz improves by up to 6 dB, but produces a pronounced resonance at 160 MHz. Ferrites in front of metal structures achieve no improvements < 30 MHz, but shows a good effect between 30 MHz and 170 MHz, and there is an increase in the emission results > 170 MHz.

?

Case 3: Comparing regular test table to a real EMC load machine (EMC-BlueBox)

The Long Wire method was used for these investigations but cannot and should not adequately replace a real component set-up and component wiring harness. Initial measurements were therefore carried out on real component set-up using the EMC-BlueBox by Frankonia with 120 kW. The results show comparability and indicate that metallic structures below the ground plane are less important. However, depending on the frequency range, the results show influences from metallic structures of a comparable magnitude, mainly due to metal structure above the ground plane. Further investigations are necessary across exiting EMC test benches for e-drives and e-axles.

?

?

Conclusions and requirements for a proper set-up and realistic measurement

Therefore, critical aspects must be considered when carrying out EMC tests:

• Normative requirements must be met as far as possible (international recognition)
• Keep metallic structures small or avoid them, especially above the ground plane
• Maximize the distance between the cable harness and metallic structures or avoid parallel routing
• It makes sense to take measurements on test objects from several directions

?

Proposal for e-axle EMC test bench

Based on CISPR 25 with an external load machine, a double arrangement with 90° angle gears is proposed, which offers various advantages:

• Metal structure within the EMC system is kept as small as possible
• Shielded shafts between external load machines into the anechoic chamber
• Shaft connection to the 90° angle gear as short as possible, but the distances according to CISPR 25 to absorbers must be observed
• Possibility of arranging the test table laterally according to CISPR 25 (focus on e-axle), but also positioned perpendicular in front of the e-axles (focus on periphery, cable harness)
• Depending on the building, load machines can also be sunk into the ground

?

?

Conclusion

The results obtained show which factors are important for an EMC test bench and which must be considered for realistic measurement. Thereof, suggestions for adapted EMC test benches for testing an e-axle are derived:

• Metallic structures in the area of the cable harness have a strong influence on the emission results
• Lining with absorbent material (ferrite) does not compensate (or only partially) for this influence and only up to 1 GHz
• Metallic structures above the table ground plane appear to be the most critical
• Due to resonance points caused by the structure, the size and position of the metallic structures above the ground plane can lead to major changes in the results. This means that measurements cannot be reproduced. Further investigations are required here.

Despite the unavoidable need to incorporate metallic structures into an EMC test site for a proper e-axle test benches, a CISPR 25 compliant test environment can be realized by considering the aspects outlined in this investigation. Higher performance classes, e.g., 2x 250 kW load machines at 3.000 RPM and 3.000 Nm (performance class based on electric axles of dominant automobile developers) have already been put into practice.

Further investigations are required. We are calling for the CISPR 25 standard to be adapted in order to enable comparable and reproducible tests on electrical axles for all OEMs and service providers internationally. Antenna positioning, dynamic driving conditions, degree of automation and a correct integration into an EMC test site also play a role. In addition to the anechoic chamber preparation, the use of the right medium for passive or active load operation of the test object (e.g., electrical, hydraulic, pneumatic) as well as the space requirements inside and outside the EMC test environment must be considered.

Frankonia Germany EMC Solutions GmbH together with EMV Babi have contributed to a practical investigation of a solution for testing electric axles, which can be implemented with various suppliers of load machines. The focus of the solution is on the correct implementation of the knowledge gained within an EMC test environment and focuses on the reproducibility of EMC tests.

The patented solution shown (EDTC-AX) offers different expansion options. More information at: www.frankonia-solutions.com

?

First publication and presentation of the results at TU Graz, September 19 to 20, 2024, 21st EMC Conference 2024, Series No. 115 of the OVE Austrian Association for Electrical Engineering.