EnglishEN
ItalianoIT
DeutschDE
EspañolES
FrançaisFR
РусскийRU
MED17 torque structure links driver demand, torque arbitration, engine models and actuator control. Understanding those relationships is more useful than memorizing a supposedly universal seven-step sequence. Map names, units and active paths depend on the engine, ECU software and definition set.
MED17 torque structure starts with the software definition
Record the ECU hardware and software identifiers, calibration version, engine and transmission before interpreting maps. A DAMOS or A2L file from another software build can contain familiar names while pointing to different addresses, dimensions or conversions. An editor label alone does not establish a map’s function.
For each relevant object, confirm its axes, units, scaling, signedness and operating conditions. A stored 16-bit value is an encoding choice, not a physical torque limit of 65,535 Nm. Likewise, percentages, normalized torque, indicated torque and torque at the clutch are not interchangeable.
Follow requests, arbitration and actuator paths
Driver demand is one input to torque coordination. Transmission requests, traction functions, operating modes and component protection can also affect the accepted request. A higher pedal-demand value therefore does not necessarily produce more cylinder filling or delivered torque.
EcuTek’s documentation for MED17.1.62 describes fast and slow torque paths: ignition and fuel interventions on one side, and air, throttle, turbocharger and cam targets on the other. That application illustrates why a simple linear map chain is incomplete; it is not a definition of every MED17 variant.
Identify which request is active and where it changes before editing a ceiling. An intervention during a shift needs different investigation from a sustained load limit or an implausible sensor signal.
Separate model maps from limits
Torque models relate cylinder filling and operating conditions to calculated torque. Inverse relationships help derive the air demand needed for a torque request. Limits constrain operation; changing a model merely to evade a limit can also corrupt the torque information used elsewhere.
| Object to identify | Question to resolve |
|---|---|
| Driver demand | Which mode selects it, and is the output normalized or in physical units? |
| Torque and air-charge models | What is the direction of conversion, reference condition and active variant? |
| Torque or load limit | When does it apply, and what component or operating constraint does it represent? |
| Monitoring threshold | Which values are compared, with what timing and plausibility conditions? |
Names such as KFMIRL or KFMIOP are useful search clues in some Bosch definitions. They do not establish universal addresses, axes or units across MED17. Do not copy a map list or percentage change from another controller without matching the actual definitions.
Investigate monitoring interventions
A torque-related fault does not prove that the monitoring threshold is too low. Preserve the fault record and available freeze-frame data, then compare demand, calculated torque, actual air charge, interventions and relevant sensor signals around the event.
Check for incorrect definitions, inconsistent models, hardware faults and unintended operating states. There is no general requirement to raise every monitor by the same percentage as driver demand. Disabling the diagnosis or inflating a threshold removes evidence without demonstrating that the underlying behavior is correct.
Validate torque reporting and the transmission interface
The torque value reported by an ECU is normally model-derived; it is not an independent dynamometer measurement. Check its meaning before comparing it with wheel torque or a corrected engine-power estimate. Loss assumptions and measurement conditions matter.
Where the transmission uses engine torque information for control, under-reporting can alter its decisions. Review ECU and TCU requests together, including shift interventions, rather than treating the engine calibration in isolation.
- Keep an identified baseline file and the tool’s documented recovery procedure.
- Record available request, limit, air-charge, pressure, lambda, ignition and temperature channels with their units.
- Compare repeatable conditions and include part-load operation and transitions.
- Investigate unexpected intervention, combustion behavior or driveline response before extending the test envelope.
An OBD calibration read is not automatically a complete memory backup. Confirm the tool’s supported read format and recovery coverage for the exact ECU. Successful writing and a clean fault scan alone do not validate the calibration.
Perguntas frequentes
Is every MED17 torque path a fixed seven-step sequence?
No. Active requests, model variants and interventions depend on the engine and ECU software. Use the matching definitions and logs.
Should torque monitors rise by the same percentage as driver demand?
There is no universal proportional rule. Identify the monitored relationship and investigate the intervention before considering a calibration change.
Is ECU-reported torque the same as measured dyno torque?
No. ECU torque is model-derived, while a dynamometer measures at its own measurement point. Definitions and losses must be understood before comparing them.