Intermittent Pedal Correlation Derate After Pedal Replacement
2017 highway tractor, 12.8 L, 289,000 miles. Intermittent 25% derate, about once per 200 miles, always clears on key cycle. Fault codes are set for accelerator pedal sensor 1 and 2 correlation. At rest with the key on, sensor 1 reads 0.75 V and sensor 2 reads 4.25 V. During the fault, sensor 1 still reads 0.75 V while sensor 2 reads 3.90 V. The pedal assembly was replaced 8,000 miles ago with a remanufactured unit. Grabbing and flexing the harness about 6 inches from the pedal connector reproduces the 3.90 V reading about 1 in 5 attempts, and the derate lamp comes on. Battery voltage and grounds at the chassis are good; the ECM 5 V reference at the pedal measures 5.01 V.
1. Interpret the relationship between the two pedal signals and explain what the 3.90 V reading during the fault tells you about where the problem lives.
Read the worked answer
The two sensors are complementary: 0.75 V + 4.25 V = 5.00 V. The ECM continuously checks that the sum (or ratio) stays within tolerance, which is why a single sensor drifting sets a correlation code even though neither sensor is out of its absolute range.
During the fault, sensor 1 is unchanged at 0.75 V while sensor 2 drops to 3.90 V, so the sum falls to 4.65 V - a 0.35 V loss on one circuit only. That is not the signature of a failed pedal (the potentiometers share a housing and usually move together, giving a scaled error on both signals). A single-channel drop, reproduced by flexing the harness, points to a circuit problem: a chafed signal wire, a terminal with lost drag, or corrosion creating resistance that changes with vibration. The pedal is on the suspect list only after the circuit is proven good.
2. Describe how you would isolate the fault to the connector terminal, the wire, or the pedal itself, using a DVOM and basic tools.
Read the worked answer
Work from the pedal toward the ECM with the connector loaded, not just probed:
- Backprobe sensor 2 at the pedal and at the ECM. Flex the harness section by section; a voltage change at the pedal but not at the ECM implicates the wire or a connector between them, while a change at both puts the fault upstream of that point.
- Min/max record the signal while flexing - a momentary 0.3 V glitch is easy to miss live.
- Check sensor ground return voltage drop under load (spec typically under 0.05–0.1 V) and the 5 V reference; a lifted ground shifts both signals and this one is normal.
- Measure end-to-end resistance of the signal wire (under 1 ohm) and wiggle-test the terminal for drag. Look for green corrosion, pushed-back pins, or a poorly crimped splice from the earlier repair.
- Confirm by overlaying a temporary wire from pedal to ECM; if the fault stops, replace/repair the harness or terminal.
3. What must be verified before and after the repair so the ticket does not come back?
Read the worked answer
Before: confirm the fault is reproducible and that the repair targets it - a wiggle test that sets the code while watching live data is worth more than a stored code. Record both sensor voltages at rest and at full pedal, and confirm the sum stays within tolerance through the full sweep.
After: clear codes, perform a full pedal sweep while watching both signals for smoothness with no dropouts, then road test with the harness dressed in its normal clips - the same clips that may have caused the chafe. Re-check that no correlation code returns after a key cycle and a load cycle. If the harness was overlaid, document the routing so the next technician knows. Also note the previous reman pedal so a repeat failure is not mistaken for a parts-quality issue.