Power Limited on the Grade: Flow, Sensor, or Inverter?

Treat 4.5 L/min against an 8.0 L/min minimum as a real restriction - a 44% shortfall. A 9 °C rise across the motor at that flow is consistent with inadequate heat removal, so start here rather than at the electronics.

Order of checks:

1. Validate the measurement. Is the pump actually at 85% duty (commanded vs. actual), and is your flow instrument appropriate for the hose and fluid? A clamp-on reading on a soft hose is easy to get wrong.
2. Air in the loop. Air locks survive a fill-and-run; the loop may need a vacuum fill and a proper bleed per procedure. The reservoir level says nothing about air trapped at the motor jacket.
3. Mechanical restriction. Kinked or internally collapsed hose, a crushed section at a clamp, a blocked cold plate or motor jacket, a stuck thermostat/flow-control valve, or a shared loop where the front motor is stealing flow.
4. Pump condition. Worn impeller or a pump that reports duty but delivers less; check for debris and for a pump that is being speed-limited by its own temperature model.

Re-measure after each step at the same commanded duty and the same coolant temperature so the numbers are comparable.

Not believable as stated. After ten minutes at idle with a 61 °C case, the winding has no heat source capable of holding a 90 °C gradient. The winding should be converging on the case temperature, not diverging from it.

Confirm or refute by measuring the sensor, not the display:

• Disconnect the sensor at the motor or inverter connector and measure resistance at the pins, then compare to the OEM's temperature/resistance table using the case temperature you measured.
• Measure through the harness as well, so you separate the thermistor from the wiring.
• Substitute a known resistor at the connector and watch the scan-tool value. If the display follows the resistor, the sensor circuit is good and the thermistor is the problem; if it does not, the harness or the module is.

Failure modes to look for: an NTC with a chafe shorted to ground reads hot (this matches the symptom). Added series resistance from a corroded connector reads cold - that would delay derate, not cause it. Coolant or oil ingress at the connector is common on motor-mounted sensors.

A sensor that reads hot is not a harmless nuisance: it drives the thermal derate model, so it produces exactly the customer's complaint with no real overheating.

Decline, on the evidence available.

• The DTC names the component whose protection strategy tripped, not necessarily the failed part. The derate is generated by a thermal model fed by coolant temperature, flow assumptions, and winding temperature.
• Two independent abnormalities are already measured: flow is 44% short, and the winding sensor is implausible. Neither is inside the inverter.
• Replacing the inverter fixes neither, and it introduces a new part into a system with an unresolved thermal complaint - you will be back, with a used inverter billed to someone.
• Note that on many designs the motor thermistor is wired through the inverter connector. That is a reason to check the harness and connector, not a reason to replace the unit.

Authorize instead: coolant loop diagnosis and repair, sensor circuit verification, then a loaded road test on a grade with the scan tool logging derate flags, winding temperature, case temperature, and actual flow. Approve the inverter only if the derate persists after flow and sensor are proven correct.