Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
CISPR 17: Methods of Measurement of Suppression Characteristics of Passive EMC Filtering Devices
Standard test methods for measuring insertion loss, impedance, and filtering performance of passive EMC suppression components
Introduction to CISPR 17
CISPR 17, originally published in 2011, establishes uniform methods for measuring the suppression characteristics of passive Electromagnetic Compatibility (EMC) filtering devices. These include discrete filter components such as inductors and capacitors, as well as complete filter units designed to attenuate conducted electromagnetic interference. The standard covers the frequency range from 10 kHz to 30 MHz, with extensions up to 100 MHz for specific applications. Understanding CISPR 17 is essential for engineers designing power line filters, signal line filters, and EMC suppression components used in industrial, commercial, and residential equipment.
When measuring common-mode (CM) insertion loss, ensure the test fixture maintains a 50 Ω/50 μH + 5 Ω asymmetric artificial mains network (AAMN) impedance to obtain reproducible results across different laboratories.
Measurement Methods and Test Fixtures
CISPR 17 specifies three primary measurement methods: the 50 Ω/50 Ω system for symmetrical filtering devices, the 50 Ω/50 μH + 5 Ω system for asymmetrical (common-mode) measurements, and the 0.1 Ω/100 Ω system for current-compensated chokes. Each method requires careful attention to fixture design, grounding, and shielding to avoid measurement artifacts. The standard provides detailed guidance on test fixture construction, including dimensional requirements for printed circuit board layouts and component mounting arrangements.
| Measurement Method | Impedance Configuration | Typical Application | Frequency Range |
|---|---|---|---|
| 50 Ω / 50 Ω | Symmetric source/load | Differential-mode filters | 10 kHz – 30 MHz |
| 50 Ω / 50 μH + 5 Ω | Asymmetric (CM path) | Common-mode chokes | 10 kHz – 30 MHz |
| 0.1 Ω / 100 Ω | Low-source / high-load | Current-compensated chokes | 10 kHz – 100 MHz |
| Reflection/Transmission | Vector network analysis | Filter S-parameters | 9 kHz – 100 MHz |
Parasitic coupling between the input and output ports of the test fixture can significantly degrade measurement accuracy above 10 MHz. Use proper shielding septa and maintain at least 20 dB isolation between ports to ensure valid insertion loss measurements.
Engineering Design Insights for EMC Filters
Practical filter design following CISPR 17 requires balancing insertion loss performance against size, cost, and thermal constraints. Ferrite core selection plays a critical role — manganese-zinc (MnZn) ferrites offer high permeability for low-frequency CM attenuation, while nickel-zinc (NiZn) ferrites perform better at high frequencies due to their higher resistivity and lower eddy current losses. Engineers must also account for DC bias current effects, which can reduce inductor saturation current and degrade filter performance by 30-50% under full load conditions.
Self-resonant frequency (SRF) is a key parameter often overlooked. A filter capacitor’s SRF determines the upper frequency limit of effective attenuation; beyond SRF, the capacitor behaves inductively and insertion loss drops sharply. Placing multiple capacitor values in parallel (e.g., 0.1 μF + 1 nF + 100 pF) extends the effective bandwidth by shifting parallel resonances. Similarly, the inter-winding capacitance of CM chokes creates a parallel resonance that limits high-frequency performance — optimizing winding geometry and layer insulation can push this resonance above 30 MHz.
Using a three-stage filter topology (CM choke + X-capacitor + CM choke + Y-capacitors) typically achieves 60-80 dB insertion loss across the CISPR 17 measurement range, sufficient for most industrial EMC compliance requirements.
Correlation Between Laboratory and Real-World Performance
CISPR 17 measurements are performed under controlled 50 Ω impedance conditions, which often differ significantly from the actual impedance environment in which the filter operates. Power line impedance varies with load current, cable length, and switching transients. Engineers should derate laboratory-measured insertion loss by 10-20 dB when predicting real-world performance, and consider using impedance stabilization networks (ISNs) that better represent actual installation conditions. Simulation tools incorporating the CISPR 17 fixture models can bridge this gap, enabling more reliable filter designs.
Never operate EMC filters beyond their rated voltage or current — core saturation in CM chokes can reduce inductance by over 90%, causing catastrophic filter failure and complete loss of EMI suppression.
Frequently Asked Questions
Q: What is the difference between CISPR 17 insertion loss and MIL-STD-220?
A: CISPR 17 specifies multiple impedance configurations (50 Ω/50 Ω, 50 Ω/50 μH+5 Ω, 0.1 Ω/100 Ω), while MIL-STD-220 uses only 50 Ω/50 Ω. CISPR 17 is more relevant for commercial EMC compliance testing.
A: CISPR 17 specifies multiple impedance configurations (50 Ω/50 Ω, 50 Ω/50 μH+5 Ω, 0.1 Ω/100 Ω), while MIL-STD-220 uses only 50 Ω/50 Ω. CISPR 17 is more relevant for commercial EMC compliance testing.
Q: Can CISPR 17 be applied to three-phase filters?
A: Yes, but each phase-to-ground and phase-to-phase path must be measured separately. The standard provides guidance on adapting test fixtures for three-phase configurations.
A: Yes, but each phase-to-ground and phase-to-phase path must be measured separately. The standard provides guidance on adapting test fixtures for three-phase configurations.
Q: How does temperature affect CISPR 17 measurements?
A: Ferrite permeability changes with temperature — MnZn cores can lose 20-40% permeability at 100°C. Measurements should be performed at 25±5°C unless otherwise specified.
A: Ferrite permeability changes with temperature — MnZn cores can lose 20-40% permeability at 100°C. Measurements should be performed at 25±5°C unless otherwise specified.
Q: What is the reproducibility of CISPR 17 measurements across labs?
A: Typical inter-laboratory variation is ±3 dB below 10 MHz and ±5 dB above 10 MHz, primarily due to fixture parasitic differences.
A: Typical inter-laboratory variation is ±3 dB below 10 MHz and ±5 dB above 10 MHz, primarily due to fixture parasitic differences.
