Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
CISPR 16-4-2: Measurement Instrumentation Uncertainty
Uncertainty analysis for EMC measurements and compliance determination
1. Scope and Importance of Measurement Uncertainty
CISPR 16-4-2 specifies the methodology for evaluating and reporting measurement instrumentation uncertainty (MIU) for CISPR-based EMC measurements. The standard addresses a fundamental question: when an EMC measurement shows the EUT’s emission at exactly the limit line, is the equipment compliant or non-compliant? The answer depends on the measurement uncertainty — the quantitative estimate of the possible error in the measurement result.
The standard introduces the concept of U-CISPR, the expanded measurement uncertainty (k=2, 95% confidence interval) for CISPR measurements. The standard provides reference uncertainty budgets for different measurement types and specifies compliance decision rules that account for the measurement uncertainty in determining whether equipment passes or fails the emission limits.
The fundamental insight of CISPR 16-4-2 is that a measurement without uncertainty is meaningless. When a laboratory reports an emission level of 55.0 dBµV/m against a limit of 56.0 dBµV/m, the true emission level could be anywhere from 52.0 to 58.0 dBµV/m (assuming ±3 dB uncertainty). Understanding this uncertainty is critical for making correct compliance decisions and avoiding costly false passes or false failures.
2. Uncertainty Budgets and Reference Values
The standard provides reference uncertainty budgets for the most common CISPR measurement types. Each budget lists all significant sources of measurement uncertainty, their probability distributions, and their contribution to the total combined uncertainty. The standard distinguishes between Type A uncertainty (evaluated by statistical analysis of repeated measurements) and Type B uncertainty (evaluated by other means such as calibration certificates and manufacturer specifications).
| Measurement Type | Reference U-CISPR (k=2) | Key Uncertainty Contributors | Typical Range Achievable |
|---|---|---|---|
| Conducted emission (LISN method, 150 kHz–30 MHz) | 3.6 dB | Receiver amplitude (±1.0 dB), LISN impedance (±0.8 dB), Cable loss (±0.5 dB), Attenuator (±0.3 dB) | 2.5 – 4.0 dB |
| Radiated emission (OATS/SAC, 30–1000 MHz) | 5.2 dB | Antenna factor (±1.0 dB), Site imperfections (±1.5 dB), EUT positioning (±0.5 dB), Cable loss (±0.5 dB) | 4.0 – 5.5 dB |
| Radiated emission (FAR, 30–1000 MHz) | 4.8 dB | Antenna factor (±1.0 dB), Site imperfections (±1.0 dB), EUT positioning (±0.5 dB), Cable loss (±0.5 dB) | 3.5 – 5.0 dB |
| Radiated emission (1–18 GHz) | 5.5 dB | Antenna factor (±1.5 dB), Site imperfections (±1.0 dB), Distance variation (±0.5 dB), Receiver amplitude (±1.0 dB) | 4.5 – 6.0 dB |
| Disturbance power (absorbing clamp, 30–300 MHz) | 4.0 dB | Clamp transfer impedance (±1.0 dB), Clamp positioning (±1.0 dB), Cable loss (±0.5 dB), Receiver amplitude (±1.0 dB) | 3.0 – 4.5 dB |
The standard specifies that the laboratory’s calculated U-CISPR must be less than or equal to the reference U-CISPR for the measurement to be considered valid. If the laboratory’s U-CISPR exceeds the reference value, the measurement results cannot be used for compliance decisions without additional justification.
The single largest contributor to radiated emission uncertainty is typically the site imperfections (±1.0–1.5 dB). This includes deviations in NSA, ground plane reflectivity variations, and absorber performance degradation. Regular site validation and maintenance are essential to keep site-related uncertainty within the budget. Many labs find that after 5 years of operation, absorber aging increases site uncertainty beyond the reference U-CISPR, requiring refurbishment.
3. Compliance Decision Rules
CISPR 16-4-2 defines two compliance decision rules. The “stringent” rule (also called the “guard band” approach) subtracts U-CISPR from the limit to create a “compliance boundary.” If the measured emission level falls below this boundary, the equipment is compliant; if above the limit, it is non-compliant. If the measurement falls between the compliance boundary and the limit, the result is inconclusive and requires improvement of the measurement uncertainty or retesting with lower uncertainty.
The “non-stringent” rule compares the measured emission directly to the limit without subtracting U-CISPR. In this case, if the measured value exceeds the limit, the equipment is non-compliant; if below the limit, it is compliant. The non-stringent rule is the default approach used in most regulatory frameworks because it provides a clear pass/fail outcome. However, it places the burden of measurement uncertainty on the consumer (the public) rather than the manufacturer — an emission source measured at exactly the limit could actually be exceeding the limit by up to U-CISPR.
The standard also defines the exclusion bandwidth, which is the frequency interval around a narrowband emission signal within which measurement uncertainty is evaluated. For CISPR Band C/D (150 kHz–30 MHz), the exclusion bandwidth is 9 kHz; for Band E/F (30–1000 MHz), it is 120 kHz.
A practical approach for reducing measurement uncertainty is to perform multiple measurements under different conditions and average the results. For example, measuring radiated emissions at two different antenna heights around the maximum position, or at two different turntable angles around the maximum azimuth, and averaging the readings can reduce the positioning uncertainty component by 0.5–1.0 dB. The standard provides guidance on when and how to apply averaging without biasing the results.
4. Frequently Asked Questions
Q: Why does CISPR 16-4-2 specify different uncertainty values for OATS/SAC vs. FAR?
A: The FAR eliminates the ground reflection, which removes the height-scanning uncertainty component. However, FAR measurements require absorber validation across the entire chamber surface, which introduces absorber uncertainty. The net effect is a slightly lower reference U-CISPR for FAR (4.8 dB vs. 5.2 dB for OATS/SAC).
A: The FAR eliminates the ground reflection, which removes the height-scanning uncertainty component. However, FAR measurements require absorber validation across the entire chamber surface, which introduces absorber uncertainty. The net effect is a slightly lower reference U-CISPR for FAR (4.8 dB vs. 5.2 dB for OATS/SAC).
Q: Can I use a U-CISPR value lower than the reference for compliance decisions?
A: Yes, provided you can demonstrate the lower uncertainty through a validated uncertainty budget. Laboratories using premium instrumentation, well-maintained sites, and carefully controlled procedures can achieve U-CISPR values 0.5–1.0 dB below the reference, which provides a commercial advantage by reducing the guard band.
A: Yes, provided you can demonstrate the lower uncertainty through a validated uncertainty budget. Laboratories using premium instrumentation, well-maintained sites, and carefully controlled procedures can achieve U-CISPR values 0.5–1.0 dB below the reference, which provides a commercial advantage by reducing the guard band.
Q: What happens if my measurement uncertainty exceeds the reference U-CISPR?
A: The measurement results are not valid for compliance decisions. You must identify and reduce the dominant uncertainty contributions — this often involves recalibrating instruments, replacing aging cables, improving site NSA performance, or refining measurement procedures. The standard requires laboratories to maintain their U-CISPR at or below the reference value.
A: The measurement results are not valid for compliance decisions. You must identify and reduce the dominant uncertainty contributions — this often involves recalibrating instruments, replacing aging cables, improving site NSA performance, or refining measurement procedures. The standard requires laboratories to maintain their U-CISPR at or below the reference value.
Q: Is measurement uncertainty considered for immunity testing as well?
A: Yes, CISPR 16-4-2 also provides uncertainty budgets for immunity measurements, including radiated RF field generation (±3 dB typical), ESD generator calibration (±15% peak current), and EFT/surge generator calibration (±10% voltage). The compliance decision rules for immunity are analogous to those for emission, but with the guard band applied to the applied test level rather than the measured value.
A: Yes, CISPR 16-4-2 also provides uncertainty budgets for immunity measurements, including radiated RF field generation (±3 dB typical), ESD generator calibration (±15% peak current), and EFT/surge generator calibration (±10% voltage). The compliance decision rules for immunity are analogous to those for emission, but with the guard band applied to the applied test level rather than the measured value.
