The email usually arrives from a distributor or a retailer, and it reads like an accusation. A customer's radio picks up a buzz whenever your product runs. A television loses signal in the evening. A baby monitor fills with clicking noise. The message ends with the question every exporter dreads: this product is certified, so how can this be happening?
It can be happening, and the certification is not contradicted by it. A test report proves something much narrower than most people assume. The fastest way to lose control of an interference complaint is to open with "we have the report" instead of finding out what changed between the unit you tested and the unit in the customer's hands.
Three complaints that all sound like "interference"
Before anything else, work out which of the three you are dealing with, because they lead to completely different places.
| What the customer describes | What it is actually about | What decides the outcome |
|---|---|---|
| Other equipment misbehaves when your product runs | Your product's emissions | Whether the shipped configuration still matches the tested one, and how much margin the original results had |
| Your product misbehaves when other equipment runs | Your product's immunity | Immunity performance - a different set of standards (EN 55035 for multimedia equipment, the IEC 61000-4 series for the individual phenomena), not the emission report |
| A licensed radio service is being disrupted | Your transmitter against protected spectrum users | Regulator involvement, and the spectrum conditions your radio approval was granted under |
The third type is the one to take most seriously, because it is the only one where the end user can be ordered to stop using the product - see the regulator section below. The first two usually end as engineering and customer-service problems, not legal ones.
What a compliant report proves, and what it never claimed
A test report is a measurement of one sample, in one configuration, against defined limits, under defined conditions: a specified cable arrangement, specified operating modes, a specified measurement distance and detector type. It is evidence about that sample. It is not a certificate and not a declaration, and it is not a warranty that the product will be quiet in every installation it ever meets.
The FCC states this in its own rules. Section 47 CFR 15.15(c) tells the parties responsible for equipment compliance that "the limits specified in this part will not prevent harmful interference under all circumstances", and encourages them to employ the minimum field strength necessary for communications and to advise users how to resolve interference problems. The statement that 47 CFR 15.105(b) requires in the manual for a Class B device says the same thing to the end user: the limits are "designed to provide reasonable protection against harmful interference in a residential installation", followed by the sentence that settles most of these arguments - "However, there is no guarantee that interference will not occur in a particular installation."
So the productive question is never "is the report real". It is: is the product in the customer's hands still the product the report describes?
Five changes that turn a compliant product into a complaint
In practice, post-sale interference traces back to one of five changes between the tested configuration and what actually shipped. None of them requires bad intent; most are ordinary production decisions.
| Change after testing | Why it shows up as interference | What it means for your documents |
|---|---|---|
| A different mains adapter or power supply | The adapter sits between your device and the mains lead, and conducted disturbance on that lead is measured with the adapter as part of the tested unit. A cheaper substitute can lift the whole conducted profile. | The report's equipment description no longer matches the shipped product |
| A firmware update | Switching frequency, LED PWM dimming frequency, radio output power, channel plan and operating modes all live in firmware. Moving any of them moves the emission signature. | May move the radio outside the configuration it was authorised in |
| A PCB revision or a second-source component | Layout, ground return paths and clock edge rates change, and radiated emissions follow. | Whether a retest is needed depends on how deep the change went |
| A different cable or antenna | Cables are the most effective unintentional antennas in the product; length, shielding quality and connector type all matter. | Cable specifications are normally fixed in the report |
| An accessory that is no longer shipped | If compliance depended on a shielded cable or a ferrite, removing it from the box removes the compliance. 47 CFR 15.27 requires that where special accessories are needed to meet the limits, the equipment must be shipped and sold with them, and the resulting system must comply. | The shipped product is non-compliant as configured, even though the tested sample was not |
Two of these account for most of what we see. The adapter is the classic: a production substitution nobody treated as a compliance decision, and the first symptom is a conducted-disturbance complaint from a user whose radio sits on the same power strip. Firmware is the quieter one. A dimming frequency was moved to make LED flicker less visible, an AM listener three rooms away started hearing a whine, and the complaint arrives through a distributor who has already promised the customer a replacement.
There is a worse version of the firmware case. A multi-mode radio product ships with a mode that was never tested - the buyer asked for a sub-GHz option late in the project and it went into the release build without the scope being updated. That is no longer a question of whether the design is good. The product on the market contains a transmitter configuration that no report describes, and no amount of arguing about the 2.4 GHz results changes that. If a change has already happened and you are unsure what closing the gap costs, the logic in which changes force a retest applies here too.
The first five days
Interference complaints go wrong when they are answered with positions instead of information. A sequence that works:
- Get the specifics. Which device is disturbed, what exactly happens, how far apart the two products are, when it happens (continuously, or only when a motor or a dimmer runs), how many units are involved, and a photo or video of the setup. "Interference" without a victim device is not actionable; a video of a radio reacting when your product switches on is.
- Classify it using the table above. If the customer's device is the victim, the emission report is the right document. If your product is the victim, it is an immunity matter and the emission report will settle nothing.
- Diff the shipped configuration against the report's equipment description: firmware version, adapter part number, antenna, cable set, everything in the box, every radio mode enabled in the release build. This is usually where the cause is.
- Reproduce it. Put the same victim device at the same distance in the same room with a shipped unit. If it reproduces, you can fix it and prove the fix. If it does not, ask for the customer's unit back - arguing from a distance has never settled one of these.
- Read your own manual. The measures the FCC requires in the Class B statement - reorienting the receiving antenna, increasing separation, moving to an outlet on a different circuit, consulting a technician - resolve a real share of complaints in homes, and many units were never installed the way the manual assumes. If those statements are missing from your manual, that is a separate problem to fix before the next shipment.
None of this requires the laboratory that issued the report. Most of it requires someone who knows what the report actually says.
When it stops being a customer argument
Two frameworks matter if the complaint escalates beyond the buyer.
In the United States, operation of a Part 15 device is conditioned on not causing harmful interference and on accepting the interference it receives (47 CFR 15.5(b)). The operator - your customer, not you - "shall be required to cease operating the device upon notification by a Commission representative that the device is causing harmful interference", and may not resume until the condition is corrected (47 CFR 15.5(c)). In practice the FCC rarely starts with a product that annoys a neighbour; it gets involved when a licensed service is affected. But the rule structure tells you where the leverage is: the complaint becomes serious when it touches protected spectrum, not when it touches a television.
In the European Union, avoiding harmful interference is part of the Radio Equipment Directive's essential requirements - radio equipment must use spectrum efficiently so as to avoid harmful interference (Article 3(2)) - and national authorities have defined procedures for radio equipment that presents a risk (Articles 40 and 42) or is formally non-compliant (Article 43), which can end in restricting, withdrawing or recalling the product. Regulation (EU) 2019/1020, applicable since 16 July 2021, requires that an economic operator established in the EU hold or be able to obtain the declaration of conformity and the technical documentation - kept for ten years after the product is placed on the market - and ensure corrective action where there is reason to believe a product presents a risk. Measures taken in one member state become visible to the authorities in the others through the ICSMS information system.
The practical point is the same on both sides of the Atlantic: the file decides how long this takes. If you can produce the report, the declaration and a configuration match between the shipped unit and the tested one, most complaints end with evidence instead of negotiation. If you cannot, you are exposed from the first letter.
What a fix usually involves
Containment first: exchange the affected units, put the correct adapter back in the box, ship the ferrite or the shielded cable the report assumed, or pull a firmware build that undoes the change. None of that is the fix; it stops the bleeding while the engineering is done.
The fix itself is usually narrow. A filter, a ferrite, a shielding change or a frequency plan adjustment addresses the specific emission causing the problem - which is why the reproduction step matters: it tells you which emission, at which frequency, under which operating mode. These are the same root causes that make products fail emissions testing in the first place, and they respond to the same remedies. Retesting the full scope is almost never necessary; retesting the affected band in the shipped configuration normally is.
What people underestimate is the paperwork that follows. If the shipped configuration no longer matches the report, the report and the declaration of conformity describe a product you no longer sell - whatever the engineering outcome. Closing that gap is a documentation exercise, not a re-run of the project, but it has to be done deliberately rather than assumed.
The short answer
A customer reporting interference is not automatically reporting a compliance failure. A compliant product is one that met defined limits, in a defined configuration, under defined conditions - and the FCC's own rules, in 47 CFR 15.15(c), state that those limits "will not prevent harmful interference under all circumstances". What decides how the complaint ends is whether the shipped unit still matches the tested configuration - same adapter, same firmware, same antenna, same cables and accessories, same radio modes - and whether you can produce the file if a market surveillance authority or the FCC asks for it. Find the change, reproduce the symptom, fix the specific emission, and bring the documents back in line with the product you are actually shipping.
Where a laboratory comes in is the part you cannot do at your desk: reproducing the complaint setup with the actual victim equipment, comparing the shipped configuration against the original test data, and running the narrow emissions retest that shows whether the fix works. If the shipped unit has drifted from the report, we will tell you that first, because it changes the whole conversation - the same configuration discipline we apply before testing, on our testing and certification programme, is what keeps you defensible after the product ships.
If a customer has reported interference, send us the complaint as written, a description of the victim device and the setup, the shipped configuration - firmware version, adapter part number, antenna and everything in the box - and the report you hold. We will come back with whether the shipped configuration matches the tested one, which emission is the likely cause, and how narrow a retest would need to be.