The Adapter Nobody Blamed

The client came to us with a device and a legal problem, in that order.

A customer had returned a mobile product with its USB charging port blackened, partially melted, and photographed in a way that left little doubt as to the urgency. There was a demand letter. There were references to medical bills, fire damage, and the CPSC. Counsel wanted an answer before deciding whether to litigate, settle, or report.

The preliminary explanation was familiar. I had seen this pattern before. The micro USB connector is a modest assembly — four conductors, a metal shell, a plastic receptacle. When it fails thermally the visible damage can be the same whether the fault is an internal short at the power pin or a high-resistance contact interface that heats under charging current. The difference lies in the details: a true short is an abrupt event, while a degraded interface is a slow betrayal of what should be a low-resistance connection.

The device itself spoke first. Under the stereo microscope the receptacle showed green corrosion product at the pin interfaces and a filamentous residue that was consistent with an electrochemical reaction. An X-ray image confirmed a low-impedance path bridging the power pin and the ground shield. The evidence suggested liquid contamination had reached the energized connector, and that the damaged connector was the point of failure. The story, as it stood, was customer-induced contamination.

Still, we replicated it.

I am always suspicious of a story that is complete before the parts have been examined. We mounted an exemplar unit on the bench, supplied it with the client’s charger, and introduced a five percent sodium chloride solution drop by drop into the micro USB receptacle. The unit was under normal charging load. The test was austere; it was intended only to confirm whether the mechanism could reproduce the damage in a plausible way.

Within a few hours the connector temperature rose above sixty degrees Celsius. By morning it had exceeded one hundred. The first sign was olfactory: the distinctive smell of heated plastic before the first discoloration appeared. Then the receptacle darkened. Then the mouth of the port charred, matching the customer photographs more closely than one would expect from a purely qualitative comparison.

We had replication. The damage mechanism was confirmed. The connector had failed in a manner consistent with liquid contamination and electrochemical heating. The case appeared to be customer-induced.

Then one of my colleagues asked what should have been the first question.

“What does the adapter do when the resistance drops that low?”

It was a simple question, but it shifted the frame. We had been looking at the visible victim — the connector, the port, the burned plastic. We had not been looking at the source of the energy that was driving the failure.

The next test was a foldback measurement. We connected the charger’s output to a decade resistor and swept from a near short circuit up through increasing resistance values, recording current and voltage at each point. Every power adapter has a characteristic curve. A well-designed adapter will detect a fault and reduce its output current at some threshold. That threshold is a deliberate choice; it is the boundary between protecting the downstream load and avoiding nuisance trips.

We tested the client’s charger, and then we tested a dozen third-party chargers from the supply room — adapters from the major vendors, the ones that had shipped with phones, tablets, and portable speakers. The client’s charger behaved differently.

It continued to deliver current at resistance values where the competitors had already folded back. Not marginally different. The client’s unit drove power into fault conditions that other adapters had already abandoned. In the scenario suggested by the failed unit, where a contaminated connector presented a variable resistance between power and ground, this difference mattered. It meant the client’s charger would sustain the fault current and the resulting heat longer than the average adapter would.

The receptacle was ordinary. The adapter was not.

That was the first inconsistency.

The second inconsistency came from the field data.

The product had shipped in the hundreds of thousands. Service returns showed roughly ten similar cases per month. That rate was not negligible, but it was not the rate one would expect if the client’s adapter were uniquely likely to sustain a dangerous fault. If the wide foldback threshold were the distinguishing risk factor, the field should have told us so. It did not.

We went back to the returns themselves, not just the statistics. We reviewed customer photographs, service intake notes, online posts. We were looking for the adapter in the frame. Not the device, not the cable, but the charger — the white brick, the wall plug, the source of the current.

That is not a detail most people record. The adapter is the invisible partner in a charging event. It ships in the box and then recedes from the narrative. But in enough cases the customer had included it, or a service technician had noted it, or a product image had accidentally captured it.

What we found was telling: the failed devices were rarely, if ever, connected to the client’s own adapter at the time of failure. Customers were using whatever charger was convenient — the one from their phone, the one they had in a drawer, a third-party unit from another product line. The micro USB standard had become sufficiently universal that interchangeability was assumed, and that assumption was the only reason the dangerous configuration had not produced a larger field problem.

The field had not acquitted the design; it had hidden behind customer behavior. The customers were accidentally insulating themselves from the risk by using adapters with more conservative foldback characteristics than the one that had shipped with the product.

The client had been lucky. Not safe.

That observation made the room quiet.

The engineers and counsel listened as we explained the distinction. The failed port had been burned by a seriously degraded connector. That was a fact. The immediate mechanism was contamination coupled with heating. That was a fact. The connector itself was not an exotic design; it was a standard micro USB receptacle. The difference lay in the adapter’s response to the fault.

The legal team asked whether this was a reportable condition.

The honest answer was that the adapter did not lack protection. It did have a current-limiting feature. It folded back, albeit at a threshold wider than the industry norm. The failure required multiple conditions to coincide: contamination, a fault path through the connector, and an adapter that would keep driving current into that path instead of shutting down entirely. The field data suggested those conditions were not coinciding at a rate that called for regulatory escalation.

The equally honest answer was that we had identified a measurable design characteristic that increased risk in a way no one had intended. We had replicated the failure. We had a claimant with photographs and a serious demand. We were choosing how to characterize a product’s safety in the absence of a broad field trend.

The decision was to revise the adapter. The client adjusted the foldback threshold to bring the charger’s response into line with the more conservative behavior we had observed in the market. The change went into production quietly. Existing units were not recalled. The claim was settled, and no CPSC report was filed.

The matter closed on defensible grounds. It did not close on the basis that the design was without flaw.

I have returned to this case more often than most.

The engineering lesson was simple enough: the burned part is rarely the only relevant part. The visible damage had been the connector, but the source of the damage was the energy source and its protection strategy. The organizational lesson was more complicated: a convenient explanation had been available early, and it was partly true. It satisfied the immediate demand of counsel. It also risked obscuring the fact that the product’s apparent safety depended on customer choices that the company had not designed for.

That is the kind of conclusion that is easy to live with when the field data is quiet and hard to live with when it is not. In this case, the field was quiet. The quiet, in itself, was not reassurance. It was a prompt to ask a different question.

The connector was the victim. The question worth asking was about the device that no one had thought to blame.