P2225: NOx Sensor Heater Sense Circuit Intermittent (Bank 2)

Intermittent electrical faults are usually characterised by what shakes them loose. This one is better characterised by what heats and cools it, and the reason is specific to where the second nitrogen oxide sensor is mounted. It sits at the far end of the aftertreatment system, and that location does something no other sensor position on the vehicle does: it goes from close to ambient in slow traffic to several hundred degrees during a filter regeneration, deliberately, repeatedly, for the life of the vehicle. Metal that is cycled through a swing that wide expands and contracts further than metal anywhere upstream, and terminals that are cycled far enough, often enough, eventually stop holding.

That gives you a test regime rather than a guess. The vehicle keeps a record of its regeneration events, and the fault has a timestamp on it. Lining the two up is the single most productive half hour available on this code. Dropouts that cluster during or shortly after regenerations point at thermal movement at a terminal or a connector body, which tells you what to look for and roughly where. Dropouts scattered across ordinary driving with no thermal event nearby point somewhere else entirely — at a conductor that has parted internally, or at the sensor's own electronics. One question, two very different investigations, and no parts bought to find out.

The second thing worth understanding about this particular fault is that it does not heal. There is a temptation, entirely reasonable on the face of it, to clear an intermittent code and wait to see whether it means anything, and on many faults that is sound practice. It is poor practice here, because the usual mechanism is mechanical and mechanical loss of retention is a one-way process. A terminal spring that has been opened past its elastic limit by repeated expansion does not recover on a cool morning. It will make contact again — that is why the light goes out — but it makes contact through less force than it did last month, and the interval between dropouts shortens rather than lengthens. Treat the first occurrence as the beginning of a trend, not as a curiosity.

There is also a cost to each dropout that is larger than it looks, and it is worth explaining to an owner who wants to leave it. This circuit is how the controller confirms the heater did what it was told. Each time confirmation is lost, the controller cannot vouch for the sensor's readiness, and on many calibrations the sensor's data is treated as unusable until the verification completes again. A dropout is therefore not a momentary blip in a signal nobody uses; it can cost a full warm-up window of valid aftertreatment data. A fault that occurs a handful of times per week quietly removes a meaningful share of the vehicle's own emissions supervision, which is exactly what an inspection is going to ask about.

When it comes to the physical work, the useful discipline is to test at the temperature the fault happens at. A connector examined cold on a lift is being examined in the condition where it works. Where the regeneration correlation is clear, the most informative inspection is of the connector body and its seal — a housing that has become brittle or has lost its grip on its seal will show it — along with terminal retention checked by feel on each pin rather than judged by how the faces look. Where there is no thermal correlation, the work moves to the conductor itself and to flexing the harness along its run while watching live data.

Finally, a note on the part. The sensor is expensive and the fault is usually cheap, but there is one case where the sensor is the honest answer even without proving it electrically: where the lead and connector are moulded to the sensor as a single assembly and the dropout has been localised to that portion, there is nothing repairable there and the assembly is the repair. Confirm which side of the connector the fault is on before deciding, because getting that wrong in either direction is expensive.