Scope: identify both controllers
This guide concerns MED17.5 engine-management investigation with a six-speed wet-clutch DSG/02E context. Owner reports identify MED17.5 on a CCZB Scirocco, while a separate K04-conversion report states only MED17 2.0 TSI. Record the engine ECU and TCU software, gearbox code, clutch hardware, coding, fuel and calibration provenance. A shared gearbox label does not justify transferring engine definitions.
Why meaningful torque reporting matters
Volkswagen's SSP 308, printed pp. 50–51, describes engine torque as a basis for 02E clutch-pressure calculation, with adaptation to clutch friction. It also describes clutch-slip calculation from gearbox input and individual input-shaft speeds on printed pp. 42–43. These are original-system principles, not a current torque-capacity rating for a modified gearbox.
Driver demand, indicated/internal torque, loss estimates, transmitted/clutch torque, a torque limit and measured wheel output are separate quantities; state the reference and units for each channel. A flat calculated value does not prove that physical torque has stopped increasing, and matching a display to one dyno point does not establish correct reporting across the range.
DQ250 torque-model changes
In a June 2022 owner/forum report on DQ250 behaviour, a MED17 2.0 TSI K04 car showed part-load torque-monitor faults and occasional slip after changes to the torque model. The success account in the same source concerns a Simos 18 car, so it cannot validate a MED17 file. Respondents favoured coherent reporting and warned that changing normalisation can affect other torque-dependent functions.
The evidence does not establish one safe MDNORM value, a universal percentage increase or a common internal TCU ceiling. Statements about relative versus absolute torque need to be interpreted against the exact signal definitions and controller implementation. The manufacturer relationship above supports investigating torque reporting; it does not prove that every slip complaint is a software fault.
A shift request and its engine response are different evidence
A torque-reduction/upshift forum report covers MED9.1 and a Simos 8.5 case. A later MED17.5.x question has no demonstrated calibration-only solution, and the map/codeword suggestions offered are not a verified MED17 recipe for changing shift intervention.
Bosch's historical ME-Motronic Technical Instruction, 1999, pp. 42–43, explains the general distinction between slower charge control and faster ignition/injection action. It predates MED17 and does not establish that family's selectors, CAN layout or minimum ignition settings.
| Complaint | Useful comparison |
|---|---|
| Boost falls during a shift. | Time-align the TCU torque request and shift phase with engine torque request, throttle, ignition, injection/cylinder state and wastegate/recirculation commands where available. |
| Slip is reported during steady loaded operation. | Compare the selected clutch/gear, input and shaft speeds, requested slip, pressure/control current and temperature. Distinguish the established gear from intentional clutch handover. |
| Behaviour changes between D, S or manual selection. | Match rpm, load, gear, temperature and shift phase before attributing the difference to a mode-specific table. |
| A torque monitor faults after model changes. | Preserve the fault and freeze-frame data, identify the changed model quantities and examine the point where the model ceases to agree with the physical response. |
Build a matched ECU/TCU record
- Retain the exact engine and transmission files and definitions used for each comparison. Confirm which requested reductions are exposed by the logger and whether ECU/TCU samples are synchronised.
- Capture pedal, rpm, load, pressure request/actual, throttle, ignition/intervention, final lambda request/actual and rail pressure alongside available gearbox requests, gear/clutch states, speeds, pressure and temperatures.
- Locate the first changing request or state. A dropped airflow setpoint is an engine-side result; it does not identify a raw CAN request or prove which controller initiated the event.
- Trace a model change through idle, part load, transitions and loaded operation as well as the peak-torque point. A model that is algebraically consistent with its inverse can still misrepresent the real engine.
- Validate any eventual engine/TCU change together. Preserve protective intervention while determining its cause rather than increasing pressure or masking slip merely to improve one run.
This is an editorial diagnostic framework, not a clutch-pressure target or a normalisation recipe. Mechanical condition, correct service state and the fitted components' limits remain part of the investigation.
Not established
- A safe MDNORM value, percentage increase or TCU torque ceiling for MED17.
- A calibration-only method of changing MED17.5.x shift intervention.
Sources and related pages
Sources: Volkswagen SSP 308, Bosch ME-Motronic Technical Instruction (1999) and the NefMoto forum reports linked above, reviewed 17 November 2026. See the CCZB injection-window follow-ups and the separately scoped MED9.1/DSG guide.
Related gearbox tuning research: DQ250 clutch adaptation, pressure and torque limits and DQ500 shift selection and gearbox identification. Each article identifies its source and vehicle scope.