Pre-Compliance EMC Testing

Pre-Compliance EMC Testing: How to Avoid Surprises

Formal EMC testing is expensive and schedules can be tight. When a product fails emissions or immunity, fixes often involve layout changes, enclosure changes, and another test booking. Pre-compliance testing helps you find the dominant issues early, when changes are still fast and inexpensive.

Most failures come from a few common mechanisms: clock harmonics, switching regulator noise, and common-mode currents on cables. The goal of pre-compliance work is repeatable, diagnostic measurements that point you to the source and let you confirm that a change actually helped. Absolute accuracy can wait for the accredited chamber.

You do not need a full chamber to get useful results. With a modest bench setup and a disciplined workflow, you can usually identify the main offenders and make progress before you pay for formal lab time.

Even short traces become effective radiators at high frequencies because radiation scales with current, length, and frequency. For a Hertzian (electrically short) dipole in the far field, the magnitude follows $|E| = \frac{Z_0 \cdot I \cdot l \cdot \sin(\theta)}{2 \lambda r}$, where $Z_0$ is the impedance of free space (377 ohm), $I$ is the current, $l$ is the conductor length, $\lambda$ is the wavelength, and $r$ is the distance. The wavelength term is what makes high frequency edges so much more efficient at radiating: at 500 MHz, a 3 cm trace on your PCB is a real emission source.

Pre-compliance testing covers both emissions and immunity, though most teams focus on emissions first because the measurements are simpler to set up. Radiated emissions are the electromagnetic fields your device puts out; conducted emissions are the noise it pushes back onto the power lines. The conducted band is typically 150 kHz to 30 MHz for CISPR 22, CISPR 32, and CISPR 14; lighting equipment under CISPR 15 is the one whose limits extend down to 9 kHz. The split at 30 MHz is somewhat artificial because conducted noise below 30 MHz often shows up as radiated emissions, with the power cable acting as the antenna. But the standards treat them separately, so you have to measure both.

Setting up your test environment is harder than it sounds. Formal testing happens in calibrated semi-anechoic chambers or on open area test sites (OATS), but pre-compliance testing usually has to make do with whatever space you have. At minimum you need a spectrum analyzer with antennas for radiated measurements and a Line Impedance Stabilization Network (LISN) for conducted measurements. The trick is knowing what your environment does to the measurements. Ambient noise from broadcast transmitters, nearby electronics, and even fluorescent lighting can bury your device's emissions. Measure the ambient first, use near-field probes to get close to the source, and do comparative tests (before/after a change) to extract useful data even in a noisy room.

Near-field probes are probably the most useful tool in your pre-compliance kit. Far-field measurements tell you pass or fail. Near-field probing tells you *why*. H-field (magnetic) probes pick up current loops and are great for finding power distribution and grounding problems. E-field (electric) probes respond to voltage variations and are better for high-impedance circuits and shielding gaps. Scan your PCB systematically while watching the spectrum analyzer and you can build an emission map that shows exactly where to focus your fixes. The near-field to far-field relationship is not always straightforward, but a strong near-field source almost always contributes to far-field problems.

Conducted emissions testing is easy to get wrong if you don't pay attention to setup and grounding. The LISN does three things: it provides a defined impedance between your device and the power source, blocks external noise from the power line, and couples the device's noise to your measurement receiver. The standard LISN impedance of 50 ohm || (50 uH + 5 ohm) approximates a typical power distribution network. Common mistakes: grounding the LISN poorly, routing cables carelessly, and forgetting that your auxiliary equipment (monitors, computers) adds its own noise. Try to replicate the intended installation as closely as you can. Cable types, lengths, and routing all affect what you measure.

For radiated emissions, you can use simpler antennas and closer distances than the formal standards require. The FCC Part 15 Class B compliance distance is 3 m, while CISPR 22/32 originally specified 10 m (3 m measurements are now also accepted with the appropriate limit-line conversion). Moving closer improves your signal-to-noise ratio, which matters in an unshielded room. The challenge is correlating what you measure with what the compliance lab will see. Antenna factors convert the voltage at your spectrum analyzer input to field strength at the antenna. Distance scaling behaves differently in the near-field versus far-field, with the boundary roughly at $r \sim \lambda/2\pi$ for electrically small sources (and larger for physically large radiators). In the far-field, field strength drops roughly as 1/r, so you can extrapolate from a closer measurement to the compliance distance, as long as the assumptions hold.

The same failure modes show up again and again, and pre-compliance testing can catch all of them. Clock harmonics usually dominate the radiated emissions spectrum: sharp spikes at integer multiples of the fundamental, often caused by poor decoupling, bad stackup, or sloppy component placement. Switching power supplies put noise on both conducted and radiated measurements, concentrated at the switching frequency and its harmonics. And common-mode currents on cables are often what actually determines pass or fail for radiated emissions. Knowing these patterns tells you where to point your probes first.

When you find emissions above limits at specific frequencies, you need to trace them back to the source. Correlating time-domain and frequency-domain measurements helps a lot. If you're seeing energy at 350 MHz, look for signals with roughly 1 ns edges. Try temporary changes during testing: add a ferrite bead, reroute a cable, insert a common-mode choke, and see what moves. The point is to understand *why* a fix works, because otherwise you're treating symptoms and the problem will resurface somewhere else.

Immunity testing is a different challenge. Emissions testing only needs passive measurement equipment, but immunity testing requires you to actively stress the device. You need signal generators and sometimes amplifiers. For basic pre-compliance, a portable ESD simulator and an EFT burst generator cover the most common tests. Radiated immunity needs an RF signal generator and amplifier, which gets expensive. Most teams limit pre-compliance immunity testing to the tests most likely to cause problems. But even something as simple as injecting common-mode noise onto cables with a current probe can reveal susceptibility issues before you pay for formal testing.

PCB Design Feedback

Pre-compliance testing shows you what's working in your PCB design and what isn't. Good stackup with solid ground planes? Lower emissions. Noisy circuits placed near sensitive ones? You'll see it in the scan.

Try experiments: add bypass caps temporarily, modify ground connections, change cable routing. Measure before and after. This iterative optimization during development is far cheaper than discovering problems during formal testing.

Cables often determine whether you pass or fail, so test them during pre-compliance. They act as efficient antennas for both transmitting and receiving electromagnetic energy. Common-mode currents (flowing in the same direction on all conductors) radiate far more than differential-mode signals. Measure common-mode current with a current probe while changing cable routing or termination and you'll quickly see where the vulnerabilities are. Try ferrite cores, different shield terminations, or connector filtering and measure the difference. How cables couple to your PCB ground structure and enclosure drives the choice of mitigation.

If your product has a shield or enclosure, test it. Perfect shielding does not exist. Seams, display cutouts, connector openings, and ventilation holes all leak. Compare emissions with and without enclosure sections installed to find the leakage paths. Near-field probe around seams and apertures to find the specific problem areas. You want adequate shielding, not maximum shielding. Overdoing it adds cost and weight and can make thermal management harder.

Document everything. Photograph your test setup, record equipment settings, note environmental conditions. When you eventually get formal compliance results, compare them to your pre-compliance data. The differences reveal systematic biases in your setup that you can correct for next time. Over time, you'll develop correction factors that make your bench measurements a reliable predictor of what the lab will show. This turns pre-compliance testing from rough debugging into a tool you can actually trust.

Pre-compliance testing saves more than just retest fees. Finding EMC issues early lets you optimize your mitigation. You might discover that a filter you planned isn't needed, or that a simpler fix works. When you understand the actual emission sources and coupling paths, you can target the root cause instead of throwing filters at everything. It also helps with design reuse: if you've validated that a circuit block is clean in one product, you have confidence it will behave in the next one.

If you're preparing for EMC certification, whether that's setting up pre-compliance testing, troubleshooting emissions problems, or recovering from a failed lab test, I'd be happy to help. I've debugged countless EMC issues and can help you find and fix problems before they become expensive.

The difference between first-pass success and multiple certification attempts often comes down to understanding where emissions actually originate and which fixes address root causes rather than just symptoms. Reach out if you'd like to discuss your situation. EMC problems are usually solvable once you know what you're looking for.

Pre-compliance testing and EMC-aware design are part of every engagement. Learn more about my pre-compliance EMC testing services to see how we help products pass certification the first time.