Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
IEC 61967-4:2002: Integrated Circuits — EM Emission Measurement (150 kHz to 1 GHz)
💡 Key Insight: IEC 61967-4:2002 is part of the IEC 61967 series that addresses electromagnetic emission measurement of integrated circuits at the component level. Part 4 specifically specifies the conducted emission measurement method using 1 Ω/150 Ω direct coupling networks. This standard is critical because IC-level EMC compliance is the foundation of system-level EMC — you cannot fix at the board level what was designed poorly at the chip level.
1. Scope and Measurement Principle
IEC 61967-4:2002 specifies a method for measuring conducted electromagnetic emissions from integrated circuits over the frequency range 150 kHz to 1 GHz. The standard applies to all types of ICs — digital, analogue, mixed-signal, and RF. The key principle is the use of 1 Ω and 150 Ω coupling networks placed in the supply lines of the IC under test. The 1 Ω probe measures RF currents on the VDD/VSS supply pins, while the 150 Ω probe measures RF voltages on I/O pins, mimicking the characteristic impedance of typical PCB trace and cable assemblies.
The standard defines a specific test board layout, including the decoupling network placement, the IC socket configuration, and the 50 Ω RF output interface to the measurement receiver (spectrum analyzer or EMI receiver). The test board is a 2-layer or 4-layer PCB designed to minimize parasitic coupling between the IC’s emission paths and the measurement instrumentation. Critical dimensions specified include the distance between the IC and the coupling network (≤ 20 mm), the width of the supply traces (≥ 1.0 mm for VDD, ≥ 1.0 mm for VSS), and the grounding via spacing (≤ 5 mm pitch around the IC perimeter).
⚠> Critical Setup Note: The most common cause of invalid IEC 61967-4 measurements is improper RF bypassing on the test board. The standard requires specific bypass capacitor values: 10 µF (electrolytic or tantalum), 100 nF (X7R ceramic), and 10 nF (NP0/C0G ceramic) in parallel at the power entry point of the test board. Missing the 10 nF NP0 capacitor is the single most frequent error — its low ESR at VHF frequencies (30–300 MHz) is essential for accurate emission measurement above 100 MHz. Substituting with X7R at 10 nF introduces measurement error of 3–6 dB above 200 MHz due to the capacitor’s self-resonance shift.
2. Test Methods and Coupling Networks
2.1 The 1 Ω Probe (Supply Pin Measurement)
The 1 Ω probe consists of a 1 Ω ± 1% resistor inserted in series with the VDD supply pin, with the voltage drop across this resistor measured via a 50 Ω coaxial connection to the measurement receiver. The probe measures RF currents flowing into the IC through the power supply pin, expressed as voltage in dBµV (which directly corresponds to dBµA into 1 Ω). The 1 Ω value is chosen as a compromise: low enough not to significantly affect the IC’s supply voltage (voltage drop at 100 mA = 100 mV), yet high enough to provide a measurable signal above the noise floor of a typical spectrum analyzer (noise floor ≈ −120 dBm at 10 kHz RBW, corresponding to approximately 0.2 µA detectable current).
2.2 The 150 Ω Probe (I/O Pin Measurement)
The 150 Ω probe is applied to I/O pins configured as outputs during the measurement. The 150 Ω impedance represents the characteristic impedance of a typical unshielded twisted-pair cable or PCB differential trace. The probe network consists of a 150 Ω resistor to ground, followed by a 6 dB attenuator (47 Ω series, 47 Ω to ground) that provides impedance matching to the 50 Ω measurement receiver input. The overall insertion loss of the 150 Ω probe is 15.5 dB, which must be compensated in the measurement results. The standard provides calibration procedures for both probes, including frequency response verification up to 1 GHz.
| Parameter | 1 Ω Probe (Supply) | 150 Ω Probe (I/O) |
|---|---|---|
| Measurement quantity | RF current (dBµA) | RF voltage (dBµV) |
| Frequency range | 150 kHz – 1 GHz | 150 kHz – 1 GHz |
| Series impedance | 1 Ω ± 1% | 150 Ω ± 1% |
| Insertion loss (probe only) | < 0.5 dB | 15.5 dB (nominal) |
| Applied to | VDD/VSS pins | I/O output pins |
| Detectable current | ~0.2 µA (typical) | ~0.03 µV (typical) |
| Maximum DC offset | 100 mV at 100 mA | N/A (AC coupled) |
3. Test Board Design and Validation
The standard mandates rigorous validation of the test board before any IC measurement. The validation procedure measures the empty test board’s ambient noise floor, which must be at least 6 dB below the expected emission levels of the IC. The board’s isolation between adjacent measurement channels must exceed 30 dB up to 1 GHz. The standard also specifies a reference IC (a known-emission device) for periodic verification of the measurement setup. The recommended calibration interval is 12 months, with a functional verification check before each measurement campaign using a comb generator or tracking generator source.
✅ Engineering Best Practice: To achieve repeatable IEC 61967-4 measurements, implement a four-point probe technique for the 1 Ω resistor connection. Rather than relying on PCB trace resistance (which introduces temperature-dependent errors of 5–15%), use separate force and sense traces from the 1 Ω resistor to the SMA connector. This compensates for the resistance of the PCB traces themselves and ensures the measurement accuracy remains within ±2% over the −40 °C to +125 °C temperature range.
4. Interpretation of Results and Correlation to System-Level EMC
The conducted emission limits in IEC 61967-4 are not pass/fail thresholds — they are measurement results that the IC manufacturer declares and the system integrator uses for board-level EMC prediction. Practical correlation between IC-level conducted emission (measured per IEC 61967-4) and system-level radiated emission (measured per CISPR 25 or CISPR 32) has been extensively studied. A general rule established by the automotive EMC community: an IC with supply pin RF current exceeding 45 dBµA at any frequency between 1 MHz and 400 MHz will likely cause a system-level radiated emission failure without additional board-level shielding or filtering.
The 2017 corrigendum (IEC 61967-4:2002/Cor 1:2017) corrected an error in the original standard’s schematic diagram of the 150 Ω probe. The original diagram showed an incorrect connection of the 47 Ω shunt resistor in the probe’s attenuator network. While this error did not affect properly built probes (which used the correct schematic from the manufacturer’s implementation), it caused confusion for laboratories constructing their own probes from the standard’s schematic. The corrigendum aligned the schematic with the actual implementation used in all commercially available 150 Ω probes.
🚨 Measurement Pitfall: The frequency range extension to 1 GHz requires careful attention to the probe’s ground return path. At frequencies above 500 MHz, a ground via inductance of just 1 nH presents 3 Ω of impedance — three times the 1 Ω measurement resistor. This parasitic impedance introduces a measurement error that increases at 20 dB/decade above 500 MHz. To mitigate this, the standard’s recommended ground via configuration (four parallel vias connecting the probe ground pad to the ground plane) should be strictly followed. Single-via ground returns routinely introduce 3–5 dB error above 800 MHz.
5. Frequently Asked Questions
Q1: What is the difference between IEC 61967-4 and IEC 61967-2?
IEC 61967-2 measures radiated emissions from an IC using a TEM or GTEM cell. IEC 61967-4 measures conducted emissions on supply and I/O pins using 1 Ω/150 Ω probes. Part 2 is a radiated measurement; Part 4 is a conducted measurement. Both are necessary for a complete IC EMC characterization.
Q2: Can IEC 61967-4 be applied to RF ICs operating above 1 GHz?
The standard’s defined frequency range is 150 kHz to 1 GHz. For RF ICs operating above 1 GHz, IEC 61967-4 measurements are still useful for characterizing baseband and control interface conducted emissions below 1 GHz. Above 1 GHz, refer to IEC 61967-6 (magnetic probe method) or IC-specific RF emission standards.
Q3: How does the test board design affect measurement reproducibility?
Test board design is the single largest contributor to inter-laboratory measurement variation. Studies show that identical ICs measured on different test boards (designed per the same standard) can yield results varying by ±6 dB. The primary causes are PCB stack-up variations, decoupling capacitor placement differences, and ground plane return path impedance variations. Using a standardized reference test board design (available from IEC or national metrology institutes) reduces this variation to ±2 dB.
Q4: Is the standard mandatory for automotive ICs?
While not mandatory in a regulatory sense, IEC 61967-4 is universally required by automotive IC procurement specifications. The AEC-Q100 Component Technical Committee and most automotive tier-1 suppliers mandatorily require IEC 61967-4 conducted emission data as part of IC EMC qualification. It is also referenced by the International Special Committee on Radio Interference (CISPR) standards for automotive receivers.
