Four separate things can limit the result
In VAG ME7.1/ME7.5 examples, a high-load calibration can be limited by the physical sensor, conversion and model, the representation of an input used by a function, its stored axis and interpolation, or another downstream limit. A graph whose axis extends further does not prove that every input and calculation can reach that range. Identify the ECU software and the particular function before applying any conclusion.
The ME7Logger alias list distinguishes rl, labelled as an 8-bit engine-load value, from rl_w, labelled as the word value. A word-sized log reaching a high number does not by itself show which representation a given ignition lookup actually uses. Check the active code path and the matched definition, including signedness, conversion and any saturation.
The 191.25% example is conditional
In the KFZW/2 axis discussion, a contributor describes an 8-bit load representation capped at 191.25% and proposes a wider-input patch. An unsigned byte with 0.75 percentage point per count illustrates the arithmetic: 255 × 0.75 = 191.25%. That result depends on the particular representation and scale; it is not a universal ME7 limit or a safe load target.
The original poster explicitly said their patch was theoretical and untested on their car. A later contributor reported that a B5 RS4 551F ME7.1 example already uses a 16-bit load axis. These statements contradict any blanket assumption that all ME7.1 files use the same width or that all ME7.5 files solve all range limits. The wiki has not independently verified the cited binaries or reproduced the patch. Do not copy code, addresses or definitions between software versions on the strength of the family name.
Check the path and the trace
- Establish the actual MAF and pressure-sensor range and transfer; compare modelled and measured values with matching units. The separate 5120 pressure-scaling guide covers a different pressure representation issue.
- For each relevant function, verify the runtime input, width, scale and any cap; then verify the axis data type, address, order and breakpoints, table dimensions and interpolation. A definition's labels alone cannot prove what the ECU executes.
- Check both cam-dependent ignition paths where fitted, their torque-model references and maps using shared axes. Extending an axis requires reviewing the output at existing points as well as the new edge.
- If an emulator or editor trace appears stuck in one cell, compare independently logged inputs, test tracing on another known table, and verify the actual image or emulator contents and the definition. Changing only an editor's axis orientation is different from changing the ECU data.
In a KFKHFM tracing case, the owner reported that a trace stayed in one cell even after returning to a stock file; another map appeared to trace. Participants suggested definition and axis checks, but the thread does not establish a final cause. Treat a trace as evidence to validate, not proof that the ECU necessarily used that cell.
See ME7 map axes and torque-model consistency and ME7 ignition and knock reference. The numerical calculation is an illustration of the cited encoding, not a recommendation to extend a map.
For checking the physical sensor transfer before investigating ECU range limits, see ME7 pressure sensor scaling and validation.