P2135: Throttle position channels disagree after normalization

P2135 detects a correlation problem between the two throttle position sensors in the reviewed GM strategy. Both channels may remain inside their individual electrical ranges while disagreeing about throttle position, so this code asks a different question from a sensor-low or sensor-high diagnosis.

Two comparisons are involved

The first comparison uses displaced TPS1 and TPS2 positions. Its allowed difference grows linearly from 6.797% at minimum throttle to 9.720% at maximum throttle. A single fixed disagreement limit therefore does not describe the complete monitor.

The second comparison evaluates normalized minimum signals against a 5% reference-voltage difference. Normalization matters: directly subtracting two raw sensor voltages does not necessarily reproduce the controller's calculation.

Confirm that correlation testing was available

Run/crank voltage must exceed 6.41 V. Individual throttle-sensor faults P0122, P0123, P0222 and P0223 must be absent, along with the P06A3 reference fault. A circuit-range fault may prevent correlation monitoring, so its presence changes which evidence should be resolved first.

The source counts either 79 failures out of 159 or 58 continuous failures at 3.125-millisecond intervals. This allows intermittent disagreement to matter even when two later readings appear consistent.

Verification should compare the appropriately interpreted channels throughout the relevant range and confirm completion of the correlation diagnostic. The monitor does not identify which channel is wrong, prove mechanical throttle sticking or authorize replacing a specific part. Exact signal interpretation requires the vehicle's electronic-throttle service information.

Source and applicability

The thresholds and component mapping apply only to the documented monitor, pending vehicle and calibration confirmation. GM diagnostic table, PDF page 601.