P22F9: A slow NOx rise can fail by initial delay or slower subsequent response

P22F9 describes slow low-to-high response of bank 1 NOx sensor 1. The cited GM monitor examines the engine-out sensor during a rising-NOx event associated with increased load. A steady concentration reading cannot by itself reproduce this dynamic assessment.

Initial delay and subsequent rise are different measures

One branch starts timing when modeled NOx reaches 30 ppm and stops when the sensor's raw reading reaches the same level. The documented delay limit is more than two seconds. Another branch compares measured and model-based rise timing, using a relative difference threshold and first-order filtering.

The second branch's definition is not completely consistent: the model-timer description ends by naming the raw signal rather than clearly identifying the model endpoint. This article does not silently correct that wording or claim a complete numerical reproduction of the relative-rise calculation.

Trustworthy inputs must precede a response judgment

The sensor must have reached its dewpoint condition, communication and electrical checks must be valid, and relevant fuel, air and EGR model inputs must not be faulted. Invalid reference information can undermine the comparison even when a raw NOx value is available.

The source allows multiple tests per trip and limits the dynamic observation to eight seconds before producing a result. Preserve the model trace, raw trace and eligibility together. The monitor supports a scoped slow-response interpretation, not an automatic sensor replacement or a universal road-test maneuver.

Source and applicability

2025 GM 04A documented monitor only. Vehicle, engine and controller calibration applicability must be established separately; no model fitment is inferred from the file. GM PDF, physical page 1167; GM PDF, physical page 1168.