Broadband Radiated Emissions Caused by a Brushed DC Motor
Every EMC engineer eventually encounters the same kind of problem: a product fails radiated emissions testing, the spectrum appears chaotic, and the first explanation seems convincing until the data proves otherwise. In one recent case, a product showed significant broadband radiated emissions across a large portion of the measured frequency range. At first glance, the spacing between several peaks suggested a digital clock source. The trace appeared to resemble a harmonic series. After a more careful investigation, however, that explanation no longer held up.
The Failed Configuration
The initial radiated emissions scan showed elevated energy across much of the band rather than a few isolated peaks. This distinction matters. A narrow set of evenly spaced emissions often points toward a clock, oscillator, switching frequency, or digital bus. A wideband signature, on the other hand, usually indicates a different mechanism: fast transient currents, poor cable return paths, motor drive noise, switching edges, grounding weaknesses, or common-mode currents flowing on external cables.
In this case, the product did not simply fail by a narrow margin at one frequency. The overall emission profile showed broad regions of elevated energy, which immediately suggested that the source was not limited to a single clean harmonic mechanism. The spectrum had structure, but it did not behave like a textbook clock problem.
Looking at the plot carefully, several observations stand out. Between roughly 50 MHz and 80 MHz, the emission level rises sharply and reaches the mid-60 dBµV/m range. This is not the behavior of a narrow harmonic spike. It is wideband energy spread over a relatively broad portion of the spectrum.
From approximately 150 MHz to 300 MHz, the trace shows another broad region where the average noise floor remains elevated. This is a frequency range where wiring, motor leads, harnesses, and internal cable structures can begin to behave like efficient unintended antennas. At these frequencies, even conductors that appear physically short from a mechanical point of view can radiate effectively when they are driven by fast transient common-mode currents.
The quasi-peak results are also scattered rather than forming a clean harmonic ladder. That is an important diagnostic clue. True clock harmonics usually exhibit consistent spacing and a reasonably predictable decay with frequency. Here, the energy is irregular, distributed, and broadband. That strongly points toward impulsive noise rather than a simple digital clock source.
Another concern is that the original trace rides close to the regulatory limit across large portions of the band. A product that barely passes, or repeatedly approaches the limit, does not have a robust compliance margin. Small variations in production, cable placement, grounding contact, enclosure fit, firmware mode, or component tolerance could push another unit above the limit during formal radiated emissions testing.
Why the Clock Hypothesis Was Misleading
The first instinct in many radiated emissions failures is to look for a clock. That is understandable. Digital clocks, microcontrollers, DC-DC converters, and high-speed interfaces often produce emissions at predictable intervals. When the spacing between peaks appears somewhat regular, it is tempting to assume that the dominant source has already been identified.
But EMC troubleshooting rarely rewards assumptions. A real harmonic problem normally has a clear fingerprint: stable spacing, identifiable fundamental frequency, and a relationship to a known oscillator, bus, or switching regulator. In this case, the peaks were not clean enough, the broadband floor was too elevated, and the emissions covered too much of the band to support a simple clock explanation.
The more likely mechanism was broadband impulsive noise coupling onto cables and radiating as common-mode current. This type of failure is common in products with motors, relays, switching power stages, long internal harnesses, poorly controlled return paths, or fast current transitions. The actual radiator is often not the noisy circuit itself but the cable or structure that provides an unintended antenna path.
What the Spectrum Was Really Saying
A radiated emissions plot should not be read only as a pass/fail result. It is also a diagnostic map. The shape of the trace, the width of the peaks, the behavior of the noise floor, and the relationship between quasi-peak and average results all provide useful clues about the source of the problem.
In this case, the broad increase between 50 MHz and 80 MHz suggested strong common-mode excitation. The elevated region from 150 MHz to 300 MHz suggested that connected wiring or internal leads were participating in radiation. The absence of a clean harmonic pattern weakened the clock-source theory. Taken together, these observations pointed toward a system-level coupling problem rather than a single noisy oscillator.
This is why pre-compliance testing is so valuable. It allows engineers to investigate the emission mechanism while the design is still flexible. Cable routing, grounding, filtering, shielding, ferrites, bonding, PCB layout, and firmware operating modes can still be adjusted before the product reaches formal certification.
Common Causes of Broadband Radiated Emissions
Broadband radiated emissions are often caused by fast transient events rather than stable periodic signals. The source may be a switching device, a motor driver, a relay, an inverter, a DC-DC converter, an ESD-like discharge, or a poorly controlled return current path. Once that energy couples onto a cable, enclosure seam, or harness, the product can radiate over a wide frequency range.
Fast switching edges from power electronics or digital circuits.
Common-mode currents flowing on motor leads, cables, or harnesses.
Insufficient grounding or bonding between circuit boards, chassis, and cable shields.
Poor separation between noisy power circuits and sensitive or externally connected conductors.
Unfiltered I/O lines, long return loops, or uncontrolled cable exits.
Enclosure seams, apertures, or mechanical gaps that reduce shielding effectiveness.
The important point is that the emission source and the radiation structure are not always the same. A switching circuit may generate the noise, but a cable may be the part that actually radiates it. This distinction is central to effective EMC testing and troubleshooting.
What This Case Teaches
This case reinforces several practical lessons that are easy to overlook during the pressure of a failed EMC test.
Not every pattern that looks periodic is caused by a digital clock.
Broadband energy usually points toward impulsive noise, switching events, or common-mode current.
Cables and harnesses often become the unintended antennas in radiated emissions failures.
A product that rides close to the limit is not robust, even if one sample technically passes.
The shape of the spectrum matters as much as the highest peak.
Good EMC troubleshooting starts with evidence, not assumptions.
How to Investigate This Type of Failure
A structured troubleshooting process is essential. The first step is to identify whether the emissions are linked to a stable clock, a switching frequency, a motor operating mode, a cable configuration, or a transient event. This can often be done by changing operating modes one at a time and observing how the spectrum responds.
Near-field probing can help locate noisy regions on the PCB, but it should not be treated as the final answer. A near-field probe may identify where energy is present, while the radiated emissions test shows how that energy escapes the product. Both views are needed. Engineers should also test cable movement, shield bonding, ferrite placement, grounding changes, load conditions, firmware states, and enclosure configurations.
For broadband issues, temporary mitigation experiments are often very useful. Adding a ferrite, improving a ground bond, changing cable routing, applying temporary shielding, or adding common-mode filtering can reveal the dominant coupling path. Once the mechanism is understood, the temporary fix can be converted into a production-ready design change.
Why Compliance Margin Matters
Passing the limit by a small amount is not the same as having a reliable design. In production, small variations are unavoidable. Cable placement changes. Component tolerances shift. Enclosure contact resistance changes. Firmware versions evolve. Operators configure products differently. Any of these variables can affect radiated emissions.
A healthy margin gives the manufacturer confidence that the product will remain compliant across normal production and use conditions. When a trace sits close to the regulatory limit across a wide band, the design deserves attention even if the prototype is only marginally failing. In many cases, improving the emission margin early is far less expensive than redesigning the product after certification failure.
How Stancer Testing-Lab Supports EMC Troubleshooting
At Stancer Testing-Lab, a very well-equipped EMC and RF Testing-Lab, radiated emissions failures are not treated as simple pass/fail events. They are engineering problems that require careful interpretation of the spectrum, the product architecture, the operating mode, and the physical test configuration. Our team supports manufacturers through pre-compliance testing, formal radiated emissions testing, and practical EMC troubleshooting.
By combining chamber measurements, engineering analysis, near-field investigation, cable and grounding experiments, and regulatory knowledge, Stancer helps manufacturers identify the real cause of emissions problems and move toward a design that is not only compliant, but robust.
Frequently Asked Questions
What causes broadband radiated emissions?
Broadband radiated emissions are commonly caused by fast transient currents, switching power circuits, motor drivers, relays, poor grounding, or common-mode currents on cables. Unlike narrow clock harmonics, broadband emissions spread energy across a wider frequency range.
How can you tell the difference between clock harmonics and broadband noise?
Clock harmonics usually show consistent spacing and predictable decay from a known fundamental frequency. Broadband noise is more irregular, often raises the average noise floor, and may appear over wide portions of the spectrum rather than as clean narrow peaks.
Why do cables radiate during EMC testing?
Cables can radiate when common-mode currents flow on them. At certain frequencies, even relatively short conductors can behave like efficient antennas, especially when driven by fast switching currents or poorly controlled return paths.
Why is quasi-peak measurement important?
Quasi-peak detectors weight emissions based on repetition rate and amplitude. They are important because many EMC standards use quasi-peak limits to reflect the interference potential of repetitive noise sources.
Why is a small compliance margin risky?
A product that sits close to the regulatory limit may pass in one configuration but fail when production tolerances, cable routing, grounding, firmware, or operating conditions change. A stronger margin improves confidence across real-world use and manufacturing variation.
How does pre-compliance testing help reduce EMC redesign costs?
Pre-compliance testing identifies emissions problems while the design is still flexible. Engineers can adjust grounding, filtering, cable routing, shielding, PCB layout, and firmware modes before formal certification testing, reducing the risk of expensive late-stage redesign.
Can Stancer Testing-Lab help troubleshoot radiated emissions failures?
Yes. Stancer Testing-Lab is an EMC/RF Testing-Lab that supports manufacturers with radiated emissions testing, pre-compliance investigations, EMC troubleshooting, near-field analysis, grounding and cable experiments, and practical recommendations to help improve compliance margins.
