Understanding the real challenges behind unstable EMC results, and why Measurement Uncertainty matter
Measurement uncertainty is one of the most misunderstood parts of EMC testing. Many teams expect emissions and immunity results to behave like simple, stable numbers. In reality, EMC measurements are influenced by dozens of variables that interact with each other in unpredictable ways. When these effects accumulate, uncertainty becomes large. This is why two accredited laboratories can test the same device and still report values several decibels apart.
This article explains why EMC measurement uncertainty is high, what drives it, and how you can reduce your risks during certification. Whether you are working on your first product launch or supporting a complex platform with many configurations, understanding uncertainty helps you design with margin, select better test partners, and make smarter decisions in your compliance roadmap.
Why EMC Measurement Uncertainty Is High
Measurement uncertainty is not a single factor. It comes from many contributions that add together. Below we explore the main sources.
1. Test Environment Limitations
Semi anechoic chambers, reverberation rooms and OATS sites all have inherent imperfections. Absorbers do not behave ideally across the full frequency range. Reflections, hot spots, partial standing waves and ground interactions all introduce variability.
Humidity, temperature and absorber aging shift the performance of the chamber day after day. With radiated emissions, these small changes easily move results by several decibels.
2. Antennas, Probes and Cables
Antennas and probes come with calibration tolerances. Cable losses change with movement, connector torque and temperature. Preamplifiers, EMI receivers, LISNs and clamps all have their own uncertainties.
These individual tolerances do not cancel out. They accumulate. And it only takes a few tenths of a decibel per component to reach several decibels in the final combined uncertainty.
3. Measurement Uncertainty, Equipment and Calibration Equipment Imparity
This part is often underestimated, yet it has a major effect on EMC repeatability.
Different laboratories use equipment from different manufacturers, with different calibration chains, software versions and reference standards. Even if everything is accredited under ISO IEC 17025, the traceability path and the calibration laboratory’s own uncertainty can differ.
Here are the main reasons why calibration imparity increases uncertainty:
- Different calibration labs
Not all calibration laboratories use the same reference antennas, same cable standards or same measurement procedures. Their own uncertainties propagate into your EMC equipment. - Divergent calibration intervals and drift
An EMI receiver calibrated 11 months ago and one calibrated last week do not behave exactly the same. Drift over time is normal, especially in preamplifiers, LISNs and field probes. - Imparity in transfer standards
A lab might use a different antenna factor reference or a different field probe calibration method. When you combine these differences with chamber imperfections, results naturally drift. - Variation in software correction tables
Receiver correction tables, antenna factors and cable loss profiles are not always aligned between labs. Even minor differences shift the measured value in borderline cases.
This imparity explains why you can run the same test at two accredited facilities and still see differences of 3 to 5 dB. Both labs can be correct. They are simply working with different calibration chains and different equipment behavior.
For manufacturers, the implication is simple. If your design passes with only 1 or 2 dB of margin, calibration imparity alone can push you outside the limits.
4. Equipment Tolerances of the DUT
The device under test generates its own variability. Switching power supplies behave differently depending on load and temperature. Digital circuits change emissions with firmware processing. Motors create fluctuating noise depending on torque.
Even units from the same batch do not behave identically.
5. Setup and Operator Influence
Cable orientation, grounding, table placement, connector tension and DUT positioning all influence the measurement. And even though standards describe these steps clearly, human adjustments introduce small deviations that add up.
Radiated emissions are extremely sensitive to cable routing. A few centimeters can change a peak by two or three decibels.
6. Statistical Behavior of EMC Phenomena
EMC emissions and immunity reactions are not deterministic. Noise sources fluctuate. Resonances move. Failures during immunity can appear at slightly different field strengths each time.
Because of this stochastic nature, laboratories must apply uncertainty models according to ISO IEC 17025.
7. Flexibility in EMC Standards
CISPR, IEC and EN standards define procedures, but they also allow acceptable alternatives. Cable types, clamp positions, table materials and measurement distances often have options.
Two labs using different but allowed interpretations produce different results.
Why This Matters for Compliance
Understanding measurement uncertainty helps you avoid design decisions that put your product at risk. A device that passes with a small margin can still fail during certification when the lab applies its uncertainty budget or when a different facility performs the test.
The safest approach is to design with margin, test early, and treat EMC results as ranges instead of absolute truths.
How to Reduce Your Risk When Facing EMC Uncertainty
Before the list, remember that you cannot eliminate uncertainty. You can only reduce its impact and plan your development around it.
Practical actions to protect your project:
- Ask the lab for their declared measurement uncertainties for each test
- Select labs with a consistent calibration chain and transparent methods
- Design with at least 3 to 6 dB EMC margin where possible
- Test early in pre compliance setups to catch resonances and sensitive cables
- Repeat critical tests on multiple units
- Keep detailed photos of every setup
- Review the lab’s equipment calibration dates before running formal tests
These actions give you a stronger position when working with certification bodies and allow you to make better decisions during NPI.
Useful Tips
- Treat EMC margins below 2 dB as high risk.
- Never rely on a single unit. Variability of the DUT itself often exceeds 1 dB.
- For immunity tests, log events with software, not with the human eye.
- Record the exact cable paths during pre compliance testing for future repetition.
