CISPR 25 Compliance Testing for Electric Vehicle Powertrains: What Engineers Need to Know

Electric vehicle powertrains are fundamentally different electromagnetic environments than the internal combustion engine systems they replace. Traction inverters switching at tens of kilohertz, DC-DC converters stepping 400V or 800V bus voltage down to 12V, and onboard chargers rectifying AC mains power—each of these subsystems generates broadband electromagnetic interference across a frequency range that spans six decades. CISPR 25, the international standard governing component-level emissions testing for vehicles, is the gatekeeper that every one of these components must pass before it reaches a production vehicle.

For engineers working on EV powertrain components, understanding CISPR 25 isn't optional—it's the difference between a smooth program launch and costly redesign cycles that can delay vehicle platforms by months.

What Is CISPR 25 and What Does It Cover?

CISPR 25 (formally IEC CISPR 25, currently in its 5th edition) is the international standard that defines methods for measuring radio disturbance characteristics of components and modules intended for use in vehicles, boats, and devices with internal combustion engines or electric propulsion. It is a component-level standard—meaning it tests individual modules on a bench, not the complete vehicle.

The standard covers electromagnetic emissions across a broad frequency range: 150 kHz to 5.9 GHz (as of the current 5th edition, extended from 2.5 GHz in Edition 4). This range is divided into conducted emissions (150 kHz–30 MHz), measured through the device's power and signal lines, and radiated emissions (30 MHz–5.9 GHz), measured with antennas in a shielded test environment. The upper frequency limit of 5.9 GHz reflects the need to protect modern vehicle receiver systems including cellular (LTE/5G), Wi-Fi, Bluetooth, GPS, and V2X communication operating above 1 GHz.

How Does CISPR 25 Differ from Other Automotive EMC Standards?

CISPR 25 occupies a specific position in the automotive EMC standards hierarchy. CISPR 12 measures emissions at the complete vehicle level—the entire car on a test site—while CISPR 25 tests individual components in isolation. ISO 11452 covers electromagnetic immunity (susceptibility), the inverse problem of whether external fields disrupt the component's operation. SAE J1113 provides similar component-level EMC test procedures used primarily in North American OEM specifications, often referencing CISPR 25 methods.

For EV powertrain suppliers, CISPR 25 is typically the first EMC gate in the development process. Failing it means the component cannot be integrated into the vehicle for CISPR 12 vehicle-level validation.

What Are the Five CISPR 25 Limit Classes?

CISPR 25 defines five limit classes (Class 1 through Class 5) with progressively stricter emission limits. Class 1 is the most permissive; Class 5 is the most restrictive and provides the greatest protection margin for onboard receivers.

In practice, nearly all major OEMs—BMW, Volkswagen Group, General Motors, Hyundai-Kia, Stellantis—mandate Class 5 for powertrain components. The rationale is straightforward: a traction inverter switching at 150 kW sits within a meter of FM, DAB, satellite radio, cellular, and GPS antennas. Class 5 limits ensure adequate margin so that conducted and radiated emissions from the powertrain do not degrade receiver sensitivity.

OEMs may also impose additional requirements beyond the base CISPR 25 standard. Volkswagen's VW TL 81000, BMW's GS 95024, and General Motors' GMW3097 all reference CISPR 25 methods but add OEM-specific frequency ranges, limit adjustments, or additional test configurations.

What Are the Key CISPR 25 Test Methods for EV Powertrain Components?

CISPR 25 specifies distinct test methods for conducted and radiated emissions, each with specific equipment, setup requirements, and measurement procedures.

How Are Conducted Emissions Measured?

Conducted emissions testing measures RF noise that the device under test (DUT) injects onto its power supply lines and signal harnesses. The measurement uses a Line Impedance Stabilization Network (LISN)—specifically the AN (artificial network) or AAN (asymmetric artificial network) types defined in CISPR 25—to present a defined impedance to the DUT and extract the noise signal for measurement by an EMI receiver.

The conducted emissions frequency range spans 150 kHz to 30 MHz. Measurements use peak, quasi-peak, and average detectors at specified bandwidths. The LISN isolates the DUT's emissions from the power supply impedance, providing repeatable measurements regardless of the specific power source used.

For EV powertrain components, the critical difference is voltage rating. Standard automotive LISNs are designed for 12V or 48V systems. A traction inverter operating from a 400V or 800V DC bus requires a high-voltage LISN (HV LISN) rated for the full bus voltage plus transient margin. Manufacturers including Schwarzbeck (NNHV 8123 series) and ETS-Lindgren produce HV LISNs specifically for EV powertrain testing.

How Are Radiated Emissions Measured?

Radiated emissions testing uses the ALSE (Absorber-Lined Shielded Enclosure) method, which is the primary method specified in CISPR 25 for radiated disturbance measurements. The DUT is placed on a ground plane inside a shielded chamber lined with RF absorbing material to suppress reflections.

Antennas measure emissions across 30 MHz to 5.9 GHz in defined sub-bands:

• 30 MHz–200 MHz: Biconical antenna (vertical polarization, 1 m distance)
• 200 MHz–5.9 GHz: Log-periodic or horn antenna (vertical polarization, 1 m distance)

The antenna is positioned at 1 meter from the DUT harness, scanned along its length to find the maximum emission point. Both peak and quasi-peak detector measurements are required at specified resolution bandwidths.

For EV powertrain testing, the harness configuration is critical. High-voltage cables carrying hundreds of amps of switched current act as efficient antennas. The harness length, routing on the ground plane, and termination method directly affect measured radiated emissions—often by 10-20 dB. CISPR 25 specifies harness length and routing requirements, but OEMs frequently add stricter constraints.

What Are the Most Common CISPR 25 Failure Points for EV Powertrains?

EV powertrain components fail CISPR 25 testing at predictable frequencies and for identifiable root causes. Understanding these failure patterns is the first step toward avoiding them.

Why Do SiC and GaN Switching Transients Cause Broadband Failures?

Silicon carbide (SiC) MOSFETs from Wolfspeed, Infineon, and STMicroelectronics switch at voltage slew rates of 10-100+ V/ns—an order of magnitude faster than the 1-5 V/ns of legacy silicon IGBTs. Gallium nitride (GaN) HEMTs in onboard chargers switch even faster, at 50-200+ V/ns. These fast transitions generate broadband noise with significant spectral content extending above 200 MHz for SiC and above 500 MHz for GaN.

The result: EV components routinely fail CISPR 25 radiated emissions limits in the 100 MHz–500 MHz range where legacy ICE components had comfortable margin. The broadband nature of the emissions makes them particularly difficult to address—unlike a discrete harmonic spur, broadband noise requires wideband filtering and shielding solutions.

How Do Inadequate Input Filters Cause Conducted Emissions Failures?

DC-DC converters and onboard chargers frequently fail conducted emissions between 150 kHz and 10 MHz due to insufficient input filtering. The switching fundamental and its low-order harmonics fall directly in the conducted emissions measurement band. A 100 kHz switching converter produces its 2nd through 300th harmonics within the CISPR 25 conducted range.

Common design errors include undersized common-mode chokes that saturate under DC bias current, Y-capacitors with excessive ESL that lose effectiveness above a few MHz, and single-stage filter topologies that cannot achieve the 60-80 dB attenuation needed for Class 5 compliance.

What Role Do Cable Shielding and Connector Bonding Play?

Poor cable shielding and connector bonding on high-voltage harnesses are the leading cause of radiated emissions failures between 30 MHz and 200 MHz. A shield braid with 85% optical coverage—common in cost-optimized designs—provides only 30-40 dB of shielding effectiveness at 100 MHz. Class 5 limits typically require 60+ dB of total system attenuation in this range.

Connector bonding is equally critical. A pigtail ground connection at the connector—where the shield braid is gathered into a single wire and soldered to a ground pin—introduces inductance that defeats the shield above a few MHz. 360-degree circumferential bonding using backshell clamps from Amphenol, TE Connectivity, or APTIV is essential for maintaining shield integrity through the CISPR 25 frequency range.

How Do Ground Loops Between the Battery Simulator and DUT Cause Measurement Errors?

The test setup itself creates failure modes. The traction battery simulator used to power the DUT during CISPR 25 testing introduces ground path impedances that don't exist in the actual vehicle. Ground loops between the simulator chassis, the ground plane, and the DUT enclosure create common-mode current paths that can increase measured emissions by 10-20 dB.

Proper bonding of the battery simulator to the CISPR 25 ground plane—using wide copper braid straps less than 200 mm long—and careful management of return current paths are essential for valid measurements.

What Strategies Help Pass CISPR 25 on the First Attempt?

First-pass CISPR 25 compliance requires a systematic approach starting early in development, not a last-minute filter and shielding retrofit.

Why Is Pre-Compliance Scanning Essential?

Pre-compliance scanning with near-field probes and a spectrum analyzer during the prototype phase is the single most cost-effective EMC strategy. A near-field probe set (H-field loops and E-field probes from Langer EMV, Beehive Electronics, or Rohde & Schwarz) costs under $2,000 and identifies emission hotspots in minutes.

By scanning the DUT before committing to formal CISPR 25 lab time—which typically costs $2,000-5,000 per day—engineers can identify and address major emission sources during development. Pre-compliance scanning won't predict exact pass/fail margins (the measurement correlation to CISPR 25 limits is approximate), but it reliably identifies whether emissions are 20 dB over the limit or 5 dB under.

How Should Filters Be Designed for CISPR 25 Frequency Breakpoints?

CISPR 25's structure creates natural frequency breakpoints where the test method, detector, or limit level changes. Effective filter design targets these breakpoints:

• 150 kHz: Start of conducted emissions band. Differential-mode LC filters with iron powder or sendust cores.
• 530 kHz–1.8 MHz: AM broadcast band. OEMs are especially sensitive to emissions here. Common-mode chokes with nanocrystalline cores (Vacuumschmelze VITROPERM) provide high impedance.
• 30 MHz: Transition from conducted to radiated measurements. Both the filter and the cable shield must be effective at this boundary.
• 76 MHz–108 MHz: FM broadcast band. Clock harmonics from gate drivers frequently land here. Spread-spectrum clocking or targeted notch filtering may be needed.
• 174 MHz–240 MHz: DAB (Digital Audio Broadcasting) band. SiC switching harmonics commonly overlap this range.

Design filters to achieve at least 10 dB of margin below Class 5 limits across all bands. Margin accounts for production variation, temperature drift, and the difference between pre-compliance and formal test setups.

What Cable Shielding Performance Is Required?

Cable shielding effectiveness for CISPR 25 compliance is characterized by transfer impedance—the ratio of voltage induced on the shield's inner surface to the current flowing on the outer surface. Lower transfer impedance means better shielding.

For EV powertrain harnesses, target:

• Transfer impedance < 50 mΩ/m from 1 MHz to 200 MHz
• Braid optical coverage > 95% with supplemental foil wrap
• 360-degree connector termination at both ends of every shielded cable
• Minimum 100 mm separation from low-voltage signal harnesses

How Should the Test Setup Be Optimized?

Aligning with the test laboratory before the first CISPR 25 session saves time and prevents invalid measurements:

• Confirm harness length and routing on the ground plane per the OEM's specification (not just base CISPR 25)
• Verify HV LISN type and impedance matches the OEM requirement (AN vs. AAN, impedance value)
• Agree on load conditions—which operating modes to test, motor load simulation method, and battery simulator grounding
• Document OEM-specific deviations from base CISPR 25—extended frequency ranges, additional measurement points, custom limit lines

Many OEMs publish their own EMC specifications that reference CISPR 25 but modify the test procedures. Volkswagen's VW TL 81000, BMW's GS 95024, and Ford's EMC-CS-2009.1 (successor to ES-XW7T-1A278-AC) all contain deviations that can cause a component passing base CISPR 25 to fail the OEM-specific version.

Frequently Asked Questions

What is the difference between CISPR 25 and CISPR 12?

CISPR 25 measures electromagnetic emissions at the component and module level, testing individual devices like inverters or DC-DC converters on a bench in a shielded chamber. CISPR 12 measures emissions at the complete vehicle level, testing the entire car on an outdoor or semi-anechoic test site. Component suppliers must pass CISPR 25; vehicle manufacturers validate the integrated system against CISPR 12.

Which CISPR 25 limit class do most OEMs require?

Most major OEMs—including BMW, Volkswagen, General Motors, and Hyundai—mandate CISPR 25 Class 5 for powertrain components, which is the most restrictive limit class. Class 5 provides the greatest margin against interference with vehicle receivers including FM radio, DAB, cellular, GPS, and Wi-Fi systems.

Why is CISPR 25 testing harder for EV components than ICE components?

EV powertrain components use high-voltage power electronics switching at high frequencies with fast slew rates (SiC at 10-100+ V/ns, GaN at 50-200+ V/ns), generating broadband EMI from 150 kHz to well above 1 GHz. Legacy ICE components like alternators and ignition systems produce narrowband disturbances concentrated below 30 MHz, making them easier to filter and shield.

What is an HV LISN and why is it needed for EV testing?

A high-voltage LISN (Line Impedance Stabilization Network) is a specialized version of the standard CISPR 25 AN or AAN network rated for 400V or 800V DC bus voltages. Standard 12V/48V LISNs cannot safely handle EV powertrain voltage levels. HV LISNs from manufacturers like Schwarzbeck and ETS-Lindgren provide the defined impedance and isolation required to measure conducted emissions on high-voltage power lines.

How can engineers reduce the risk of failing CISPR 25 testing?

Pre-compliance scanning with near-field probes and a spectrum analyzer during development is the most cost-effective strategy. This identifies emission hotspots before committing to formal lab time. Additionally, designing input and output filters targeting CISPR 25 frequency breakpoints (150 kHz, 30 MHz), maintaining cable shielding transfer impedance below 50 mΩ/m, and engaging the test lab early to align on OEM-specific test deviations all significantly improve first-pass yield.

Why Early Investment in CISPR 25 Compliance Pays Off

CISPR 25 compliance is non-negotiable for EV component suppliers. A single test failure triggers a redesign cycle—new filter components, modified PCB layout, revised shielding—followed by another round of lab time. The typical cost of a CISPR 25 failure recovery is 8-16 weeks of schedule delay and $50,000-200,000 in engineering and retest costs.

The components that pass CISPR 25 on the first attempt share common traits: pre-compliance scanning started during schematic review, filter design budgets included 10 dB of margin, cable shielding specifications included transfer impedance requirements, and the engineering team aligned with the test laboratory on OEM-specific procedures before the first test session.

Investing in EMC design rigor during the development phase—including proper shielding design, filter optimization, and pre-compliance validation—is consistently more cost-effective than addressing CISPR 25 failures after the fact.

---

Need EMI shielding solutions for CISPR 25 compliance? Contact POCONS USA to discuss shielding requirements for your EV powertrain components.